Copyright TechTarget - All rights reserved ComputerWeekly’s best articles of the day https://cyber.law.harvard.edu/rss/rss.html Techtarget Feed Generator en Fri, 13 Mar 2026 17:19:14 GMT https://www.computerweekly.com [email protected] <p>Organisations are investing heavily to harness <a href="With%20research%20showing%20the%20use%20of%20AI%20may%20temporarily%20reduce%20productivity,%20Cloudera’s%20Vini%20Cardoso%20urges%20businesses%20to%20adopt%20an%20organisation-wide%20platform%20approach%20driven%20by%20measurable%20value%20and%20trusted%20data">artificial intelligence</a> (AI) to solve business problems, but putting the technology into production presents significant data quality and availability challenges.</p> <p>This is especially true for applications requiring real-time data, such as <a href="https://www.computerweekly.com/news/366621012/How-Oracle-is-using-AI-to-combat-financial-crime">fraud detection</a>, according to Confluent’s chief technology officer (CTO) Stephen Deasy.</p> <p>Speaking to Computer Weekly on the sidelines of the Confluent Data Streaming World Tour in Melbourne, Deasy noted that organisations are <a href="https://www.computerweekly.com/feature/AI-From-exploration-to-production-five-case-studies-on-GenAI-in-action">moving beyond experiments and trials</a> this year. They are taking what they have learned about AI and applying it in high-impact areas to deliver customer value, leveraging the latest streamed data to achieve their goals.</p> <p>Locally, there is a move to build on existing systems of record and engagement to create “systems of action,” which require much fresher data, added Greg Taylor, senior vice-president and Asia-Pacific general manager at Confluent.</p> <p>What’s needed is to capture data where it is created and then process it in real time so it can be used to inform action, such as to automate parts of the business, he said. Some Confluent customers in the region have been able to achieve 60 to 70% automation, although Australian organisations are generally not that advanced, he observed.</p> <p>Taylor said such systems will require business experts as part of the governance structure – for example, to <a href="https://www.techtarget.com/searchenterpriseai/tip/A-short-guide-to-managing-generative-AI-hallucinations">check for AI hallucinations</a> – but they do enable continuous improvement based on data.</p> <p>Being able to react quickly to signals can increase revenue and reduce fraud, and the ability to run the Confluent platform on-premise, in the cloud or in a hybrid environment helps customers achieve such goals.</p> <p>“AI is top of mind for customers,” said Deasy, adding that Confluent helps enterprises deliver real-time data directly to their AI models.</p> <p>The company’s adherence to open standards plus its investment in technology and support resonates with customers, he said, and the company often works with CTOs and systems architects to help accelerate the implementation process.</p> <p>Technology and performance improvements for various workloads are coming, so Confluent can deliver for what people are doing today and what they will be doing in the future, said Deasy.</p> <p>Taylor observed that customers traditionally rely on suppliers for new features and capabilities, but the big <a href="https://www.computerweekly.com/news/366546575/The-rise-of-Generative-AI-in-software-development">improvements in software engineering brought by generative AI</a> (GenAI) means they can potentially create their own features, and this gives them greater negotiating power.</p> <p>“We see that regularly,” admitted Deasy, noting that this creates constant pressure on the company’s own software engineering teams to produce more, faster.</p> <p>The Melbourne event also highlighted real-world deployments, featuring presentations from three major Australian enterprises: Bendigo Bank, Telstra and Coles.</p> <section class="section main-article-chapter" data-menu-title="Bendigo Bank"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Bendigo Bank</h2> <p>Sam Fursdon, principal AI engineer at Bendigo Bank, described how using the Confluent platform, including Confluent Flink, enabled the bank to significantly reduce the mainframe load generated by its <a href="https://www.computerweekly.com/news/252471528/Australia-gets-ready-for-open-banking">open banking obligations</a> and its mobile-only subsidiary, Up. Importantly, Confluent Flink integrated seamlessly with Bendigo’s <a href="https://www.computerweekly.com/ehandbook/A-Computer-Weekly-buyers-guide-to-continuous-integration-and-continuous-deployment">continuous integration and continuous deployment</a> (CI/CD) patterns, creating a “well-orchestrated deployment pipeline”.</p> <p>By combining transaction and balance data in Confluent, the bank reduced mainframe <a href="https://www.computerweekly.com/feature/Enterprise-strategies-for-API-management">application programming interface (API) calls</a> by 50%. The overnight batch processing backlog was cleared and is now completed by 6am.</p> <p>Furthermore, the average end-to-end latency between a transaction occurring and the information becoming available for consumption is now just 2.3 seconds during business hours. In practice, this means an ATM user can receive an app notification before the cash has even been dispensed.</p> </section> <section class="section main-article-chapter" data-menu-title="Telstra"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Telstra</h2> <p>Telstra uses Confluent to improve its mobile network and customer experience. Quoting a colleague, Javed Bolim, Telstra’s technology product owner for observability, said: “You need to see it to action it.”</p> <p>Events are captured continuously and streamed in real time for analysis. This helps with early detection of issues before the customer notices, provides richer context by correlating more signals, and the scalability of the system means new uses can be implemented without the need to re-engineer data flows.</p> <p>Data streams are filtered, enriched, and stored in a database for multiple uses. These include service assurance at major events, such as Boxing Day test matches at the Melbourne Cricket Ground (MCG), and proof of value, ensuring customers get what they pay for. Bolim added that new products and features enabled by this capability are currently in the pipeline.</p> <p>Bolim advised other IT practitioners to focus on developing the right skills among team members and securing business buy-in when difficult choices must be made. He also recommended exceeding intermediate goals and developing shared capabilities, such as providing self-service access for other internal teams.</p> </section> <section class="section main-article-chapter" data-menu-title="Coles"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Coles</h2> <p>Supermarket chain Coles faced challenges stemming from the use of dozens of disparate event-based systems across various parts of the organisation. “The stuff was everywhere,” said principal engineer Simon Bedford. This sprawl led to duplication, additional costs, operational friction and the need for a security redesign.</p> <p>Coles invested heavily in deploying Confluent as a true enterprise platform, incorporating tooling, monitoring, observability, data products and discoverability. Despite the scale of the deployment, it did not require a large team, relying on just three to five people at various stages of the project.</p> <p>The initiative provided an opportunity to start afresh and enforce architectural principles, including consistent naming conventions, clear ownership boundaries, and automated provisioning.</p> <p>Governance must be baked in and automated rather than manual, Bedford advised. He noted that the platform is treated “like an internal software-as-a-service [SaaS]”, with developers acting as the customers. Because developers will typically work around anything they find overly restrictive, it was important to get the user experience right, making education a critical part of the process. Ultimately, providing a strong developer experience, including self-service provisioning through <a href="https://www.techtarget.com/searchitoperations/definition/GitOps">GitOps</a> and CI/CD integration, resulted in strong internal adoption.</p> <p>“We put in a lot of work to make sure [observability and monitoring] are enterprise-grade,” he added, noting that cost attribution is now achieved through telemetry.</p> <p>As a result, “chaos and complexity” have been replaced with “structure and efficiency”, and improved data discoverability has led to high levels of reuse across the business.</p> <p>“The platform is now easier than the alternatives and is more reliable,” said Bedford. Because it is trusted, it is widely used, delivering business outcomes that include faster time to market, reduced integration costs and improved customer responsiveness. “We’ve got high-quality data, [and] we want to look at how we can use it for AI.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in Australia</h3> <ul class="default-list"> <li>The Australian government has struck a major&nbsp;<a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">five-year volume sourcing agreement with Microsoft</a>&nbsp;to speed up adoption of AI and cloud technologies across the public sector.</li> <li>ANZ Bank has started <a href="https://www.computerweekly.com/news/366638802/ANZ-rolls-out-AI-agents-for-business-bankers">rolling out AI agents within its new CRM system</a> to help business bankers recover hours of lost productivity by automating tasks and streamlining workflows.</li> <li>With research showing the <a href="https://www.computerweekly.com/news/366639906/What-it-takes-to-succeed-with-AI">use of AI may temporarily reduce productivity</a>, Cloudera’s Vini Cardoso urges businesses to adopt an organisation-wide platform approach driven by measurable value and trusted data.</li> <li>Wesfarmers has signed a <a href="https://www.computerweekly.com/news/366639137/Wesfarmers-to-deploy-agentic-AI-in-retail-operations">multi-year deal with Google Cloud</a> to deploy AI across its portfolio of brands, including Kmart, Officeworks, Priceline and OnePass.</li> </ul> </div> </div> </section> Moving AI from experiment to production requires high-quality, real-time data streaming. Australia tech leaders from Confluent, Bendigo Bank, Telstra, and Coles share how they are turning systems of record into systems of action https://cdn.ttgtmedia.com/visuals/German/article/data-speed-flash-storage-adobe.jpg https://www.computerweekly.com/news/366640193/Why-real-time-data-is-key-for-enterprise-AI Thu, 12 Mar 2026 23:03:00 GMT <p>Enterprises are failing to realise the full potential of <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence</a> (AI) and risk actively draining employee productivity unless they adopt an organisation-wide platform approach that brings models directly to their data.</p> <p>That’s the warning from Cloudera’s chief technology officer for Australia and New Zealand, Vini Cardoso, who called for businesses to move from <a href="https://www.computerweekly.com/news/366639344/APAC-firms-still-in-AI-test-mode-as-data-readiness-issues-persist">fragmented, ad-hoc AI experiments</a> towards empowering staff and targeting use cases that deliver measurable returns.</p> <p>In an interview with Computer Weekly, Cardoso said it is also critical to consider the <a href="https://www.computerweekly.com/feature/AI-projects-yield-little-business-value-so-far">value any AI project can bring</a> to the business. An estimate that serves to secure approval is one thing, but that value must be tracked through the implementation.</p> <p>“We’ve seen organisations, like one of the banks in the region we work with, achieve A$150m of annual value creation through AI across multiple formats – not just <a href="https://www.computerweekly.com/news/366612652/APAC-organisations-embrace-generative-AI">generative AI</a>, but traditional <a href="https://www.techtarget.com/searchenterpriseai/definition/machine-learning-ML">machine learning</a> as well. This was achieved by selecting the right use cases that drive the right value.”</p> <p>By focusing on the right initiatives – those where they could gain efficiencies, reduce risk and avoid losses – those organisations were able to achieve measurable results.</p> <p>“In any transformational use case – whether AI today or <a href="https://www.computerweekly.com/resources/Enterprise-resource-planning-ERP-software">ERP [enterprise resource planning]</a> in the past – you need business leaders on board. They have to embrace the change, which they will only do if they understand the positive outcomes. Once they do, they should lead by example, establishing a culture that supports the whole organisation towards that goal, giving teams the opportunity to learn and experiment, while keeping focus on the transformation objectives,” said Cardoso.</p> <p>This change presents IT and data professionals with an opportunity to redefine their place in the organisation, taking their technology expertise and using it to influence business decisions and outcomes.</p> <p>For example, a data analytics team might examine quarterly business reviews and provide a recommendation along the lines of: “Maybe you should invest in these opportunities, as based on our analysis they are higher value and more likely to close, rather than using a scattergun approach across many prospects.”</p> <p>“In my experience, this makes a more satisfying job. Employees are no longer just operating software: they see real outcomes and influence critical business decisions,” Cardoso said.</p> <p>However, a recent survey by Sapio Research for Foxit Software suggested that the use of AI may temporarily reduce rather than improve productivity.</p> <p>That survey of 1,400 desk workers and executives in the US and the UK found executives save just 16 minutes per week, while desk workers lose 14 minutes, due to the time needed to validate AI outputs. A likely related finding was that one in four executives, but only one in 10 desk workers, are “extremely confident” in AI outputs.</p> <p>Cardoso noted that technology adoption often starts with scepticism, but trust grows over time as value is demonstrated. “Take self-driving cars: in California, I saw ride-hailing cars operating with no one inside. The level of sophistication and safety implementation is very high, and gradually, people gain confidence and adoption increases.</p> <p>“The same is happening with AI adoption in business. Organisations take measures to increase confidence in AI outcomes, starting with trusted data: knowing where it comes from, what transformations were applied, and ensuring reliable sources,” he added.</p> <p>People must be able to trust both the model and the data it is fed. Keeping <a href="https://www.techtarget.com/searchenterpriseai/feature/Humans-and-AI-The-role-of-people-in-the-new-AI-world">humans in the loop</a> during the early stages helps develop trust, and that – along with a corresponding increase in risk appetite – accelerates adoption.</p> <p>“The human side – technical skills plus business knowledge – remains critical to review frameworks and guide decisions. And as society evolves, frameworks will need to adapt. Models must evolve too. The beauty of AI is that it learns continuously, naturally adjusting outcomes over time to reflect new knowledge and changing expectations,” Cardoso said.</p> <p>Another recurring issue is moving from experiments and pilots to production systems. Those with the potential to deliver real value are more likely to gain the necessary support, according to Cardoso.</p> <p>He also recommended adopting a platform approach from the start. “Sometimes people design solutions using whatever technology is at hand, but that technology may not be scalable or operationalised. The time required to recode, re-architect, or re-engineer data pipelines can be massive.</p> <p>“With a platform approach, you have a consistent way to tap into data sources and develop AI for production. Data and workloads become reusable and standardised. You can apply consistent security, governance and controls, making the transition from pilot to production seamless. At the same time, you adhere to governance frameworks and address regulatory and compliance requirements,” he said.</p> <p>To that end, Cloudera offers trial use of its software so potential customers can test both its capability and their own use cases.</p> <p>The platform can be deployed on-premise, in the customer’s selected cloud, across multiple clouds, or in a hybrid environment. This means organisations can choose whichever cloud is cheapest at the time while maintaining the same workload, infrastructure and governance standards. That includes the ability to place workloads to take advantage of unused pre-committed cloud credits.</p> <p>Partly due to the cost of moving data between clouds, “the approach we take is to bring the AI workloads to the data, rather than moving the data to the AI,” Cardoso said. “If the data is on-premise, we help deploy AI capabilities to run models there. If it’s in the cloud, we deploy workloads close to where the data lives.”</p> <p>However, in appropriate circumstances, for example, where there are low or no egress fees, the Cloudera platform has federation capabilities, allowing access to data across different environments without moving it.</p> <p>“The platform approach ensures AI runs where the data is, without exposing IP or sensitive information – critical for sovereignty requirements. AI should augment skills, unleash ideas and generate value, not replace teams.</p> <p>“High-value use cases are essential. Low-hanging fruit is useful for learning, but targeting real business problems drives investment and board-level buy-in for scaling AI,” Cardoso said.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in APAC</h3> <ul class="default-list"> <li>NetApp’s regional chief discusses the <a href="https://www.computerweekly.com/news/366639344/APAC-firms-still-in-AI-test-mode-as-data-readiness-issues-persist">gap between AI intent and production</a>, the rise of neoclouds, and why the storage firm is counting on getting data AI-ready to win market share.</li> <li>DayOne and Cortical Labs are <a href="https://www.computerweekly.com/news/366639849/Neurons-over-silicon-Singapore-plans-first-biological-datacentre">bringing ‘wetware’ computing to the city-state</a>, using living neurons grown from stem cells to support the demand for AI while addressing sustainability concerns.</li> <li>While firms in mature markets are using AI agents to automate routine tasks, those in emerging markets where the cost of the technology is higher than that of human labour are <a href="https://www.computerweekly.com/news/366639619/Emerging-markets-prioritise-top-line-growth-with-agentic-AI">favouring revenue-generating use cases</a>.</li> <li>The Australian government has struck a major <a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">five-year volume sourcing agreement with Microsoft</a> to speed up adoption of AI and cloud technologies across the public sector.</li> </ul> </div> </div> With research showing the use of AI may temporarily reduce productivity, Cloudera’s Vini Cardoso urges businesses to adopt an organisation-wide platform approach driven by measurable value and trusted data https://cdn.ttgtmedia.com/visuals/German/article/data-network-data-analysis-adobe.jpg https://www.computerweekly.com/news/366639906/What-it-takes-to-succeed-with-AI Wed, 11 Mar 2026 23:49:00 GMT <p>Customer service software giant Zendesk has entered into a definitive agreement to acquire Forethought as it looks to boost the capabilities of its <a href="https://www.computerweekly.com/feature/Getting-started-with-agentic-AI">artificial intelligence (AI) agents</a>, which the company expects will handle more customer interactions than human agents this year.</p> <p>Through the proposed acquisition, which Computer Weekly understands is Zendesk’s largest in two decades, Forethought’s self-improving AI will be integrated into Zendesk’s Resolution platform, allowing AI agents to generate, adapt and execute complex workflows across different service channels.</p> <p>“The era of simply managing conversations is over,” said Tom Eggemeier, Zendesk CEO. “The future of customer experience requires agentic capabilities built for definitive resolution. Forethought’s advanced capabilities perfectly align with our vision for agentic service. Together, we will be scaling self-improving AI that learns from every interaction. But technology is just the means. Resolution is our identity, and loyalty is the outcome.”</p> <p>Zendesk’s AI agents can learn from customer conversations without the need for manual retraining, working in tandem with human agents to resolve over 80% of routine interactions today. With Forethought, Zendesk will be able to provide specialised AI agents for business-to-business, consumer and employee use cases, as well as native voice automation.</p> <p>Notably, the Forethought platform will extend AI into existing enterprise systems even where <a href="https://www.computerweekly.com/feature/Enterprise-strategies-for-API-management">application programming interfaces</a>&nbsp;(APIs) do not exist, eliminating manual work and unlocking previously unreachable workflows.</p> <p>Sami Ghoche, co-founder and CEO of Forethought, noted that joining Zendesk was the fastest way to accelerate the startup’s mission to transform customer experience for every business.</p> <p>“With Zendesk’s platform, resources and global reach, we will bring our technology to many more organisations around the world, move faster on innovation and continue pushing the boundaries of what AI can do in customer experience,” he said.</p> <section class="section main-article-chapter" data-menu-title="The agentic AI acquisition wave"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The agentic AI acquisition wave</h2> <p>The proposed buyout of Forethought is the latest in an aggressive consolidation phase across the enterprise software market as suppliers race to infuse agentic AI capabilities into their platforms.</p> <p>Over 50 agentic AI acquisitions have been announced globally over the past two years, according to <a href="https://zinnov.com/automation/why-software-leaders-are-racing-to-acquire-agentic-ai-blog/">Zinnov</a>, a management consulting firm. Zendesk’s chief rival, Salesforce, has heavily invested in this area, <a href="https://www.techtarget.com/searchcustomerexperience/news/366624032/Salesforce-acquisition-of-Convergence-adds-Agentforce-talent">acquiring startups like Convergence.ai</a> and <a href="https://www.techtarget.com/searchcustomerexperience/news/366639253/Salesforces-acquisition-of-Cimulate-to-boost-AI-shopping">Cimulate</a> to power its <a href="https://www.computerweekly.com/news/366638802/ANZ-rolls-out-AI-agents-for-business-bankers">Agentforce platform</a>.</p> <p>Meanwhile, IT service management software supplier ServiceNow <a href="https://www.techtarget.com/searchitoperations/news/366620527/ServiceNow-agentic-AI-acquisitions-set-pace-for-enterprises">spent nearly $2.85bn to acquire enterprise search and AI agent developer Moveworks</a> last year, while customer experience supplier <a href="https://omdia.tech.informa.com/om138028/nices-acquisition-of-cognigy-a-new-era-for-ai-powered-cx-automation">Nice acquired Cognigy</a> for about $955m to bolster its CXone Mpower platform.</p> <p>Zendesk itself is no stranger to the M&amp;A rush. The Forethought deal follows the company’s <a href="https://www.constellationr.com/insights/news/zendesk-acquires-ultimate-builds-out-ai-agent-platform">2024 acquisition of Finnish service automation provider Ultimate</a>, which laid the groundwork for Zendesk’s current AI agent offerings. Last December, <a href="https://www.zendesk.com/sg/newsroom/articles/zendesk-acquires-unleash/">it also acquired Unleash</a>, an AI-powered enterprise search platform, to shore up its employee service capabilities.</p> <p>Keith Kirkpatrick, vice-president and research director at The Futurum Group, said that with the deal, Zendesk is making a bold statement that agentic AI will define the next era of customer experience. “At a time when many software companies are cautious or still in pilot mode, this investment reflects strong confidence in both the technology and the market’s readiness,” he added.</p> <p>Chuck Ganapathi, CEO of Gainsight, a joint customer of both Zendesk and Forethought, noted that service must be autonomous and deeply integrated to deliver a world-class customer experience.</p> <p>“For Gainsight, this deal will provide the sophisticated, cross-platform automation we need to ensure every customer interaction is intelligent, seamless and aligned with our broader mission for driving retention for our customers,” he said.</p> <p>Zendesk said the Forethought acquisition will accelerate its product roadmap by over a year. Forethought will remain available to new customers without requiring them to use the core Zendesk platform. The transaction is expected to close by the end of March, subject to regulatory approvals.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about agentic AI in APAC</h3> <ul class="default-list"> <li>Singapore has launched a <a href="https://www.computerweekly.com/news/366637674/Singapore-debuts-worlds-first-governance-framework-for-agentic-AI">governance framework agentic AI&nbsp;systems</a> to address the growing security and operational risks posed by AI agents.</li> <li>While firms in mature markets are using AI agents to automate routine tasks, those in emerging markets where the cost of the technology is higher than that of human labour are <a href="https://www.computerweekly.com/news/366639619/Emerging-markets-prioritise-top-line-growth-with-agentic-AI">favouring revenue-generating use cases</a>.</li> <li>Australian retail giant Wesfarmers has signed a <a href="https://www.computerweekly.com/news/366639137/Wesfarmers-to-deploy-agentic-AI-in-retail-operations">multi-year deal with Google Cloud</a> to deploy AI agents&nbsp;across its portfolio of brands, including Kmart, Officeworks, Priceline and OnePass.</li> <li>Microsoft is <a href="https://www.computerweekly.com/news/366636341/Microsoft-taps-Indian-IT-giants-for-agentic-AI-rollout">teaming up with four of India’s largest IT services companies</a> to deploy agentic AI capabilities across enterprises.</li> </ul> </div> </div> </section> Zendesk is acquiring Forethought to bolster its agentic AI chops with specialised and self-learning AI agents capable of managing complex customer service workflows https://cdn.ttgtmedia.com/visuals/German/article/Hero-Laptop-Customer-Service-Headset-rh2010-adobe.jpg https://www.computerweekly.com/news/366639959/Zendesk-to-acquire-Forethought-in-major-agentic-AI-play Wed, 11 Mar 2026 08:30:00 GMT <p>Imagine servers that don’t hum with the intense heat of thousands of power-hungry silicon chips but instead rely on “wetware”, biological neurons grown from stem cells that can process information and power <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence (AI)</a> workloads with just a fraction of the energy required by classical computers.</p> <p>This science-fiction scenario could soon be a reality, with DayOne, a Singapore-headquartered global datacentre provider, teaming up with <a href="https://www.computerweekly.com/news/366537095/How-lab-grown-neurons-could-power-the-future-of-AI">Melbourne-based biocomputing startup Cortical Labs</a> to build Singapore’s first biological datacentre.</p> <p>Supported by capital and strategic input from DayOne, the project will start off as a prototype at the Yong Loo Lin School of Medicine at the National University of Singapore (NUS Medicine).</p> <p>Founded by Hon Weng Chong, Cortical Labs first grabbed global headlines in 2022 with its <a href="https://www.sciencedirect.com/science/article/pii/S0896627322008066">DishBrain project</a>, where researchers successfully taught a network of 800,000 in-vitro brain cells to play the classic arcade game <em>Pong</em> within five minutes.</p> <p>What differentiates the startup’s technology – which <a href="https://www.allthingsdistributed.com/2023/03/australia-the-new-epicenter-for-healthtech-startups.html">caught the eye of Amazon chief technology officer Werner Vogels</a> – from traditional computing is its reliance on biological intelligence rather than <a href="https://www.techtarget.com/searchenterpriseai/definition/neural-network">neural networks</a>. Because biological cells naturally strive to reduce unpredictability in their environments, they can learn and adapt with incredible efficiency. More importantly, they consume very little energy and generate almost no heat.</p> <p>The prototype at NUS Medicine will comprise a single rack of 20 <a href="https://corticallabs.com/cloud">Cortical Cloud units</a>. Under the supervision of Rickie Patani, a professor and director of the neurobiology programme at the NUS Life Sciences Institute, the cells will be cultured and grown to test the system’s viability for complex research.</p> <p>“Wetware systems can help researchers explore new approaches to learning, adaptation and biological modelling,” said Patani. “Our expertise in neurobiology research, particularly in understanding how to generate specific subtypes of clinically relevant human neurons and glia from stem cells, provides a strong foundation for translating these biological principles into biocomputing platforms.</p> <p>“For applications such as <a href="https://www.computerweekly.com/news/366623123/Pharma-leaders-see-AI-revolutionising-medicine">drug discovery</a> and <a href="https://www.techtarget.com/healthtechanalytics/news/366589848/University-of-Massachusetts-DL-Model-to-Predict-Alzheimers-Onset">neurological disease research</a>, the ability to run experiments on brain-like biological networks alongside conventional computing could accelerate hypothesis testing and shorten cycles from laboratory insight to meaningful real-world impact,” he added.</p> <p>Following the validation phase at NUS, the collaboration will transition into a live deployment environment inside a DayOne commercial datacentre in Singapore. The wetware systems will be tested under real-world workloads, tracking compatibility with standard power distribution, contained environmental management systems and cooling infrastructure. If successful, the partners are exploring a phased expansion that could see up to 1,000 units deployed.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Singapore has made it clear that the next chapter of digital infrastructure must be built with sustainability at the core </figure> <figcaption> <strong>Hon Weng Chong, Cortical Labs</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>The initiative comes at a time when global datacentre capacity is projected to reach 200GW by 2030, with power demand for datacentres in Southeast Asia expected to quadruple from 2.6GW in 2025 to 10.7GW by 2035. In response, the Singapore government has enacted sustainability policies for the sector, releasing at least 200MW of new capacity under its green datacentre roadmap.</p> <p>“Singapore has made it clear that the next chapter of digital infrastructure must be built with sustainability at the core,” Chong said. “AI is moving from novelty to necessity across every sector, but the region’s energy and water realities are forcing a reckoning. This partnership is about giving policymakers and industry a practical alternative – a sustainable pathway to AI adoption that aims to decouple compute growth from a resource footprint.”</p> <p>For DayOne, wetware offers an alternative computing model for a sector grappling with power availability and grid emissions.</p> <p>“Singapore is raising the bar for sustainable datacentre growth, and the market is responding with new approaches, beyond just bigger builds,” said Jamie Khoo, CEO of DayOne. “Partnering with Cortical Labs allows us to explore a new compute paradigm that complements Singapore’s and the region’s sustainability-led trajectory, supporting continued innovation while staying aligned to evolving efficiency and greener energy expectations.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentres in APAC</h3> <ul class="default-list"> <li>Industry leaders discuss how the computational demands of advanced AI models are forcing a <a href="https://www.computerweekly.com/news/366623023/How-AI-workloads-are-reshaping-datacentre-design">rethink of datacentre power</a>, cooling and networking infrastructure.</li> <li>Digital Realty is <a href="https://www.computerweekly.com/news/366637290/Digital-Realty-enters-Malaysia-with-acquisition-of-Cyberjaya-datacentre">set to acquire CSF Advisers</a>, the owner of the TelcoHub 1 datacentre in Cyberjaya, marking its expansion into Malaysia to address the country’s fast-growing demand for digital infrastructure services.</li> <li>India has restarted consultations on its&nbsp;<a href="https://www.computerweekly.com/news/366629299/India-revives-national-datacentre-policy-amid-AI-push">long-delayed national datacentre policy</a>&nbsp;to unlock nationwide opportunities and prepare for the surge in AI and cloud workloads.</li> <li><a href="https://www.computerweekly.com/news/366629279/Macquarie-Data-Centres-to-offer-Dell-Nvidia-AI-tech-stack">Macquarie Data Centres</a>&nbsp;teams up with Dell Technologies to host the Dell AI Factory with the Nvidia infrastructure platform in its Australian facilities.</li> </ul> </div> </div> DayOne and Cortical Labs are bringing ‘wetware’ computing to the city-state, using living neurons grown from stem cells to support the demand for AI while addressing sustainability concerns https://cdn.ttgtmedia.com/visuals/German/HERO-KI-AI-Alexander-Machine-learning-stockAdobe-07.jpg https://www.computerweekly.com/news/366639849/Neurons-over-silicon-Singapore-plans-first-biological-datacentre Wed, 11 Mar 2026 00:42:00 GMT <p>Digital sovereignty has become one of those phrases everyone repeats but few define. In Asia, it is often reduced to a simple “US versus non‑US cloud” debate or a box‑ticking exercise around data localisation – which makes for good talking points, but does not help a Singapore-based chief financial officer (CFO) decide where to run a core banking system or a regional chief information security officer (CISO) determine how to train an AI model on customer data without tripping over three different regimes at once.</p> <p>Asia’s digital economy has also outgrown the habit of defining sovereignty purely in reaction to other markets: from Indonesia’s data protection law to Vietnam’s Decree 53, the region is building its own foundations for a data‑driven future, even as global <a href="https://www.computerweekly.com/opinion/AI-and-cloud-The-perfect-pair-to-scale-your-business-in-2025">cloud and artificial intelligence (AI) platforms</a> are deeply woven into how businesses operate, from supply chains to customer engagement.</p> <p>Framing sovereignty as “US versus non‑US” misses this nuance. It assumes the only choice that matters is whose flag flies over the datacentre, when most organisations in Asia will inevitably use a blend of local, regional and global providers. The real questions are far more practical and more uncomfortable: which laws can legitimately claim jurisdiction over data and systems; how easily critical workloads can be moved or re‑architected if risk changes; and who, in practice, can access the data at 2am during an incident.</p> <p>These are the conversations we need to be having. <a href="https://www.techtarget.com/searchcio/feature/Ignoring-digital-sovereignty-CIOs-cant-afford-to">Digital sovereignty</a> in Asia only becomes useful when we stop treating it as a slogan and start treating it as a design discipline built on three pillars: data sovereignty, technical sovereignty and operational sovereignty.</p> <section class="section main-article-chapter" data-menu-title="Data sovereignty: Law before location"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Data sovereignty: Law before location</h2> <p>Data sovereignty is often mistaken for simply “where the data sits”, but location is only one part of the equation. What really matters is jurisdiction: which courts and regulators can reach your data, especially as AI rules emerge across different markets. As AI spreads into areas like credit, hiring, healthcare and public services, regulators are moving beyond soft guidance – <a href="https://aibasicact.kr/">South Korea’s new AI Basic Act</a> is a clear signal that higher‑impact systems will face stricter expectations on risk, oversight and transparency. In this environment, losing control of data quickly becomes losing control over how businesses operate and innovate.</p> <p>The real issue here is clarity and choice. Boards should be asking where data is stored and processed, which legal entities control the infrastructure, what commitments exist around using data for AI training and how government access requests are handled. In Asia, data sovereignty should be about mapping different classes of data – from public content to highly sensitive personal information – to environments where the legal and contractual frameworks match the level of risk.</p> </section> <section class="section main-article-chapter" data-menu-title="Technical sovereignty: No more “stuck in one stack”"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Technical sovereignty: No more “stuck in one stack”</h2> <p>Technical sovereignty comes next: the ability to move, rebalance or redesign workloads when business, regulatory or geopolitical realities shift. The biggest sovereignty risk may not be a particular law but over‑dependence on a single cloud stack.</p> <p>No single stack will satisfy every jurisdiction as AI and data rules evolve. Measures that are voluntary in one market today can become mandatory in another tomorrow. That is why <a href="https://www.computerweekly.com/feature/Cloud-data-portability-Obstacles-to-it-and-how-to-achieve-it">data portability</a>, clear workload segmentation and the ability to shift between public, private and more sovereign environments are becoming design requirements rather than nice‑to‑haves.</p> <p>In a region as diverse and fast‑moving as Asia-Pacific, multi‑cloud and hybrid are not fashion statements. Done thoughtfully, they are tools of sovereignty, giving organisations room to manoeuvre when regulations tighten, AI policies evolve or geopolitical tensions spill over into the technology stack.</p> </section> <section class="section main-article-chapter" data-menu-title="Operational sovereignty: Where cyber security and sovereignty meet"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Operational sovereignty: Where cyber security and sovereignty meet</h2> <p>Operational sovereignty is often the missing piece in board conversations. Yet, it is where sovereignty becomes very real. Even with the right contracts and architectures, what matters day to day is who can access systems and data, from where and under what controls.</p> <p>In a globalised services model, it is normal for support engineers, site reliability teams and security specialists to sit outside the customer’s home market. The issue is not geography per se, but governance. For example: are privileged accesses strictly logged, time‑bound and justified?; which jurisdictions do the provider’s support operations sit in and what obligations they are under?; can in‑region support models for certain workloads be insisted, especially in regulated sectors?&nbsp;</p> <p>This is where cyber security and sovereignty converge. <a href="https://www.computerweekly.com/resources/Identity-and-access-management-products">Identity and access management</a>, <a href="https://www.computerweekly.com/news/366613540/Applying-IT-observability-to-deliver-business-metrics">observability</a>, incident response and audit trails are not only security topics, they are operational proof of control. After an incident, regulators and customers in Asia are increasingly asking, not just “were you secured?” but “who else could see this data, under what rules and how does one know?”. For Asian leaders, operational sovereignty should be treated as part of their cyber resilience strategy, not a separate compliance concern.</p> </section> <section class="section main-article-chapter" data-menu-title="Designing for freedom of choice"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Designing for freedom of choice</h2> <p>Ultimately, sovereignty should not be treated as a checkbox at the end of a cloud procurement process. It needs to be incorporated from the outset as a way to preserve freedom of choice over time by classifying data and workloads according to sensitivity demanding transparency from partners on jurisdictions, data flows and support models; building enough technical portability to respond if laws, risk or economics change; and treating operational controls around access and logging as part of the sovereignty posture, not just the security posture.</p> <p>Organisations across Asia-Pacific that move the fastest are not those seeking the most rigid constraints, but those striving for trusted flexibility. They want a cloud they can trust, not one that traps them. They want a cloud that provides assurance, so they know their data is protected and have the freedom to innovate, grow and adapt as the regulatory and geopolitical landscape evolves.</p> <p>It is worth remembering that sovereignty in Asia will not be won by copying someone else’s debate. It will be defined by the region’s ability to build cloud strategies that put Asian priorities at the centre: legal clarity, technical agility, operational trust and, ultimately, the freedom to chart our own digital future.</p> <p><i>Terry Maiolo is vice-president and general manager for Asia-Pacific at OVHcloud</i></p> </section> Framing digital sovereignty simply as a “US versus non-US cloud” debate is no longer fit for purpose. Asia must forge its own path through data jurisdiction, technical portability, and operational control https://cdn.ttgtmedia.com/visuals/German/Hero-Data-Sovereignty-Natalia-03.jpg https://www.computerweekly.com/opinion/Why-Asia-needs-its-own-model-of-digital-sovereignty Tue, 10 Mar 2026 04:04:00 GMT <p>Government agencies in Western Australia are putting sensitive data and critical public services at risk due to inadequate Microsoft 365 (M365) security configurations, a <a href="https://audit.wa.gov.au/reports-and-publications/reports/microsoft-365-security-controls-state-entities/">report by the state’s auditor general</a> warns.</p> <p>In its audit, the WA Office of the Auditor General (OAG) evaluated seven government agencies and found widespread vulnerabilities in governance, <a href="https://www.computerweekly.com/resources/Identity-and-access-management-products">identity and access management</a>, information protection, logging and monitoring, and threat prevention.</p> <p>WA auditor general Caroline Spencer said these vulnerabilities increase the likelihood of cyber incidents, data breaches and operational disruptions. She added that, with evolving cyber threats, effective management of M365 security is essential to safeguarding sensitive government data and ensuring key public services continue to be provided.</p> <p>The OAG decided not to identify the agencies involved to prevent threat actors from immediately targeting them, thereby reducing the risk of further security breaches. However, the report included case studies that showed how poor M365 security controls can have real-world consequences.</p> <p>In one instance, a state entity emailed the personal and sensitive information of 32 individuals, including minors, to a third-party service provider. A threat actor was able to access the data because the provider kept it in an unmanaged Dropbox account that was later compromised in a cyber security incident.</p> <p>The OAG found that the entity lacked <a href="https://www.techtarget.com/whatis/definition/data-loss-prevention-DLP">data loss prevention</a> (DLP) controls to safeguard the data or verify the full extent of the exposure. It also failed to conduct a security assessment of the third-party during vendor onboarding.</p> <p>In another incident, a senior officer’s M365 account at a state entity was compromised by a threat actor using a targeted phishing email. The threat actor bypassed current security measures, registering their own <a href="https://www.computerweekly.com/ehandbook/Multifactor-authentication-What-are-the-pros-and-cons">multi-factor authentication (MFA)</a> techniques on an unmanaged foreign device.</p> <p>The breach went undetected for a month, during which multiple suspicious activity warnings were ignored. The attacker created email forwarding rules to conceal communications, studied the officer’s email history to craft a believable scenario, eventually sending fraudulent invoices that were approved for payment and stealing A$71,000 from the entity.</p> <p>The OAG’s technical assessment found that while the audited agencies had DLP controls, they were not applied to OneDrive, SharePoint, Power Platform, Exchange and Teams, leaving these platforms vulnerable to data leaks and unauthorised access. And where policies were implemented, they did not protect all kinds of sensitive data.</p> <p>In addition, instead of deploying phishing-resistant authentication for privileged users, some agencies had relied on MFA methods, such as SMS messages, voice calls and email one-time passwords that are highly susceptible to phishing attacks. The OAG noted that compromised accounts were responsible for 39% of reported cyber incidents targeting the Australian government in 2024-25.</p> <p>The audit also revealed poor <a href="https://www.techtarget.com/searchitoperations/answer/What-logging-best-practices-can-help-with-log-management">data logging practices</a>, with some entities only retaining audit logs for six months, far short of the Australian Signals Directorate’s (ASD) recommended 18-month retention period, reducing their ability to trace the origin and full impact of security incidents.</p> <p>“Effective management of M365 security is critical for protecting sensitive government data and maintaining uninterrupted delivery of essential public services amid evolving cyber security threats,” Spencer said.&nbsp;</p> <p>“I recognise that the rapid and complex evolution of technology consistently presents new challenges for entities. To effectively counter emerging threats, entities must remain alert and continue strengthening their security posture.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about cyber security in Australia</h3> <ul class="default-list"> <li>Australian privacy commissioner warns that the <a href="https://www.computerweekly.com/news/366633983/Fewer-data-breaches-in-Australia-but-human-error-now-a-bigger-threat">human factor is a growing threat</a> as notifications caused by staff mistakes rose significantly even as total breaches declined 10% from a record high.</li> <li>Exabeam’s chief AI officer, Steve Wilson, urges organisations to&nbsp;<a href="https://www.computerweekly.com/news/366631420/Exabeam-Treat-AI-agents-as-the-new-insider-threat">treat AI agents as the new insider threat</a>&nbsp;and monitor them for rogue behaviour just like their human counterparts.</li> <li>Ping Identity CEO Andre Durand explains why&nbsp;<a href="https://www.computerweekly.com/news/366629810/How-to-secure-the-identity-perimeter-and-prepare-for-AI-agents">identity has become the critical security battleground</a>&nbsp;and how decentralised credentials will reduce data breach risks.</li> <li>Security chiefs at the&nbsp;<a href="https://www.computerweekly.com/news/366620535/How-CISOs-are-tackling-cyber-security-challenges">Gartner Security and Risk Management Summit in Sydney</a>&nbsp;share insights on navigating board communication, organisational resilience and the importance of understanding business needs.</li> </ul> </div> </div> Western Australia’s Office of the Auditor General has uncovered weaknesses in M365 configurations across seven government agencies, leading to compromised accounts and data breaches https://cdn.ttgtmedia.com/visuals/German/article/Hero-cloud-security-Alexandra-02.jpg https://www.computerweekly.com/news/366639954/WA-auditor-flags-weak-Microsoft-365-security-controls-across-state-entities Tue, 10 Mar 2026 01:39:00 GMT <p>Bitdefender researchers have uncovered an <a href="https://www.darkreading.com/cyberattacks-data-breaches/ai-python-variant-brazilian-whatsapp-attacks">artificial intelligence (AI)-assisted malware propagation</a> campaign, allegedly by a Pakistani threat group, that has been industrialising cyber attacks in South Asia.</p> <p>Leveraging an emerging malware category known as “vibeware”, the campaign has been linked, with medium confidence, to APT36, a state-sponsored threat group – also known as Transparent Tribe – that has historically been associated with targeting the Indian government, diplomatic missions and defence-related entities.</p> <p>Rather than aim for technical sophistication, the vibeware model relies on <a href="https://www.computerweekly.com/feature/LLMs-explained-A-developers-guide-to-getting-started">large language models</a> (LLMs) and AI-powered development tools to rewrite malicious logic across multiple programming languages, generating large volumes of malware variants almost daily.</p> <p>Researchers observed malware samples written in niche languages such as <a href="https://nim-lang.org/">Nim</a>, <a href="https://ziglang.org/">Zig</a> and <a href="https://crystal-lang.org/">Crystal</a>, along with more widely used languages like <a href="https://www.techtarget.com/searchapparchitecture/tip/Why-is-Rust-a-critical-programming-language">Rust</a> and <a href="https://www.techtarget.com/searchitoperations/definition/Go-programming-language">Go</a>.</p> <p>By using less commonly monitored languages, the group effectively resets the detection baseline for many traditional security tools. Bitdefender described this tactic as a form of “distributed denial of detection”.</p> <p>“Rather than a breakthrough in technical sophistication, we are seeing a transition toward AI-assisted malware industrialisation that allows the actor to flood target environments with disposable, polyglot binaries,” Bitdefender researchers noted in a <a href="https://businessinsights.bitdefender.com/apt36-nightmare-vibeware">blog post</a>.</p> <p>While the volume of malware is high, the quality of code in vibeware is often low. Bitdefender’s analysis revealed that many of the samples contained coding flaws and incomplete logic consistent with <a href="https://www.computerweekly.com/news/366639364/How-AI-code-generation-is-pushing-DevSecOps-to-machine-speed">AI-assisted code generation</a>.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about cyber security in APAC</h3> <ul class="default-list"> <li>Singapore mobilised over 100 cyber defenders to neutralise a sophisticated APT actor which infiltrated Singtel, StarHub, M1 and Simba networks in the country’s&nbsp;<a href="https://www.computerweekly.com/news/366638973/Singapore-mounts-largest-ever-cyber-operation-to-oust-APT-actor">largest coordinated cyber incident response to date</a>.</li> <li>Japan’s Nikkei has confirmed a major data breach that potentially&nbsp;<a href="https://www.computerweekly.com/news/366634243/Nikkei-data-breach-exposes-personal-data-of-over-17000-staff">exposed the personal information of more than 17,000 employees</a>&nbsp;and business partners after hackers infiltrated its internal Slack messaging platform.</li> <li>Australian privacy commissioner warns that the&nbsp;<a href="https://www.computerweekly.com/news/366633983/Fewer-data-breaches-in-Australia-but-human-error-now-a-bigger-threat">human factor is a growing threat</a>&nbsp;as notifications caused by staff mistakes rose significantly even as total breaches declined 10% from a record high.</li> <li>Philippine bank&nbsp;<a href="https://www.computerweekly.com/news/366633428/BDO-Unibank-taps-Zscaler-to-secure-cloud-migration">BDO is shoring up its cyber security capabilities</a>&nbsp;to protect its data and systems as it moves more services to the cloud and expands its physical presence into remote areas of the archipelago.</li> </ul> </div> </div> <p>In one instance, a basic Go binary was deployed to steal browser credentials, but the developers left a template placeholder where the <a href="https://www.techtarget.com/whatis/definition/command-and-control-server-CC-server">command-and-control</a>&nbsp;URL should have been, meaning the tool could never actually exfiltrate data.</p> <p>“We saw similar patterns across the rest of the fleet, where other malware components began to collapse under their own weight as soon as the logic reached a moderate level of complexity,” the researchers explained. “These kinds of mistakes are typical of code that is syntactically correct but logically unfinished.”</p> <p>Despite these basic errors, the overall strategy remains effective. The sheer volume and diversity of malware variants increase the likelihood that at least one implant will evade <a href="https://www.techtarget.com/searchsecurity/tip/How-antivirus-software-works-Virus-detection-techniques">signature-based or behaviourally tuned malware detection engines</a>.</p> <p>In several cases, victims were infected with multiple parallel implants written in different languages and using separate communication protocols. If defenders block one access path, others remain active, significantly complicating incident response and increasing operational resilience for the attackers.</p> <section class="section main-article-chapter" data-menu-title="Living off trusted services"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Living off trusted services</h2> <p>To further obfuscate detection, APT36 has been piggybacking on trusted services. Instead of relying solely on attacker-controlled infrastructure, the vibeware makes use of legitimate services such as Google Sheets to store malware instructions, or Slack to send real-time instructions or retrieve harvested data.</p> <p>This allows malicious traffic to blend seamlessly into normal business activity, making detection and disruption far more difficult. Bitdefender’s investigation into APT36’s internal infrastructure also revealed a recurring developer persona known as Nightmare, who appears to be central to the development and operation of the malware fleet.</p> <p>“While this malware lacks true technical innovation, it would be a mistake to underestimate the risk it poses,” the researchers warned. “The threat lies in the industrialisation of these attacks. We are seeing a convergence of two trends that have been developing for some time: the adoption of exotic, niche programming languages, and the abuse of trusted services to hide in legitimate network traffic.”</p> <p>While the targeting remains highly focused on South Asian regional politics and national security, the implications of AI-assisted malware assembly lines extend globally. AI is significantly lowering the barrier to entry for experimenting with new languages and delivery mechanisms, proving that even imperfect code can succeed when deployed at scale.</p> <p>For organisations across the broader Asia-Pacific region, Bitfender said the findings underscore the need for layered detection strategies that prioritise behavioural analysis, anomaly detection and monitoring of trusted cloud services, rather than relying solely on static signatures.</p> </section> The Pakistani threat group has been using AI to rewrite malicious code across multiple programming languages, prioritising scale over sophistication to evade detection, security researchers have found https://cdn.ttgtmedia.com/visuals/German/article/malware-1-adobe.jpg https://www.computerweekly.com/news/366639830/APT36-unleashes-AI-generated-vibeware-to-flood-targets Mon, 09 Mar 2026 06:30:00 GMT <p>The increasing usage of <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence (AI)</a> and data sovereignty requirements is driving the growth of neoclouds and sovereign cloud providers. While these emerging cloud providers offer computational capacity at a lower cost, the market is fragmented, with geographical restrictions and a lack of managed AI services.</p> <p>According to recent data from Gartner, worldwide spending on sovereign cloud infrastructure as a service (IaaS) is set to reach $80bn in 2026, representing a 35.6% increase from 2025. This structural shift, which Gartner terms “geopatriation”, is expected to see 20% of existing workloads move from global hyperscalers to local or regional cloud providers.</p> <p>By 2030, neocloud providers, which focus on crunching <a href="https://www.computerweekly.com/feature/Gartner-Considerations-when-using-GPUs-in-the-datacentre">graphics processing unit (GPU)-intensive AI workloads</a>, are expected to account for around 20% of the $267bn AI cloud market.</p> <p>However, extracting value from these emerging cloud suppliers requires a clear-eyed assessment, according to Adrian Wong, director analyst for cloud infrastructure and operations within the Gartner for Technical Professionals (GTP) research organisation.</p> <p>In a recent interview with Computer Weekly, Wong explained that while hyperscalers like Amazon Web Services (AWS), Google Cloud, and Microsoft Azure offer comprehensive, value-added AI services, neoclouds such as GPU specialist CoreWeave and sovereign providers like Europe’s Scaleway are geared towards infrastructure services.</p> <section class="section main-article-chapter" data-menu-title="‘HPC on steroids’"> <h2 class="section-title"><i class="icon" data-icon="1"></i>‘HPC on steroids’</h2> <p>For organisations looking to deploy high-end AI capabilities, neoclouds offer an attractive proposition by offloading incredibly complex underlying hardware requirements.</p> <p>“It’s like HPC on steroids, where the interconnection between GPUs, the network, and connection to storage needs to be highly optimised,” Wong said, referring to <a href="https://www.computerweekly.com/resources/Clustering-for-high-availability-and-HPC">high-performance computing</a>. “Otherwise, you’re starving your GPUs and you’re not maximising your return on investment. So, there’s a level of value that they provide from that perspective.”</p> <p>The raw infrastructure capabilities of neocloud providers can bring cost advantages. Many provide <a href="https://www.computerweekly.com/opinion/Gartner-Why-neoclouds-are-the-future-of-GPU-as-a-Service">cost savings of up to 60–70%</a> compared with hyperscaler GPU instances, while offering near-instant access to the latest hardware generations, according to Mike Dorosh, senior director analyst at Gartner.</p> <p>However, unlike hyperscalers like Google Cloud with <a href="https://www.techtarget.com/searchenterpriseai/tip/Compare-Google-Vertex-AI-vs-Amazon-SageMaker-vs-Azure-ML">Vertex AI and AWS with Amazon Bedrock</a>, neoclouds don’t typically offer managed AI services.</p> <p>“I can take advantage of AI capabilities from hyperscalers and use them in my organisation; I don’t necessarily need to be a data science expert,” Wong said. “It can be a lot harder when you’re starting off with a blank page. That’s the struggle that we’re seeing in the neocloud market at the moment.”</p> </section> <section class="section main-article-chapter" data-menu-title="Spotty availability"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Spotty availability</h2> <p>Despite heavy investments from neoclouds and sovereign providers, the global availability of state-of-the-art GPUs remains inconsistent outside of the US.</p> <p>Wong pointed to Nvidia-backed CoreWeave as an example. While it is one of the biggest players in the neocloud space, its GPU footprint outside the US is limited to markets such as Canada, Sweden, Spain and Norway. Similarly, European provider Scaleway has built out regions in Paris, Amsterdam and Warsaw, but GPU availability is limited to one or two datacentres within those regions.</p> <p>“They’re trying to build things out, but it’s very spotty,” said Wong. “That’s going to be very jarring for an enterprise used to dealing with a hyperscaler. If you’re looking at a neocloud or sovereign cloud provider, it’s significantly more inconsistent, and there’s a lot less documentation and transparency about what’s going to be available.”</p> <p>Hyperscalers aren’t about to cede the market to neocloud and sovereign cloud providers without a fight. For one thing, they already provide services such as <a href="https://www.computerweekly.com/news/366557158/AWS-to-open-European-sovereign-cloud-region">AWS European Sovereign Cloud</a> and <a href="https://www.computerweekly.com/news/366542413/Oracle-opens-first-EU-based-sovereign-cloud-regions-in-Germany-and-Spain">Oracle EU Sovereign Cloud</a> to meet the demand for sovereign services.</p> <p>Wong noted that sovereignty exists on a spectrum, and hyperscalers are trying to hit every point along it – from restricted cloud regions used by governments and the defence sector to on-premise deployments via services like <a href="https://www.computerweekly.com/news/366574294/Oracle-Cloud-A-discussion-about-public-cloud-dedicated-regions-and-Alloy">Oracle Cloud Infrastructure Dedicated Cloud region</a>, as well as hybrid environments via offerings such as <a href="https://www.computerweekly.com/news/252474952/AWS-Outposts-debuts-in-Australia">AWS Outposts</a> or <a href="https://www.techtarget.com/searchwindowsserver/feature/Azure-Local-aims-to-answer-shifting-needs-of-the-enterprise">Azure Local</a>.</p> <p>Ultimately, true sovereignty often comes down to geopolitics and business goals. While organisations in markets such as Australia continue to rely heavily on US hyperscalers due to historically strong ties, those in Europe or the Greater China region are increasingly evaluating sovereign cloud services out of geopolitical caution.</p> <p>Still, Wong advised that any move to a neocloud or sovereign provider must be grounded in business strategy and internal technical capabilities.</p> <p>“If I’m hoping to take advantage of a whole bunch of capabilities, and then they aren’t even offered in a neocloud or a sovereign cloud provider, that can rule things out,” Wong said. “Some specialised cloud providers, neoclouds and sovereign clouds just aren’t equipped or used to serving the needs of enterprise customers.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about cloud and AI in APAC</h3> <ul class="default-list"> <li>Australia’s Digital Transformation Agency has struck a major five-year&nbsp;<a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">volume sourcing agreement with Microsoft</a>&nbsp;to speed up adoption of&nbsp;AI and cloud&nbsp;technologies across the public sector.</li> <li>Wesfarmers has signed a&nbsp;<a href="https://www.computerweekly.com/news/366639137/Wesfarmers-to-deploy-agentic-AI-in-retail-operations">multi-year deal with Google Cloud</a>&nbsp;to deploy&nbsp;agentic AI&nbsp;across its portfolio of brands, including Kmart, Officeworks, Priceline and OnePass.</li> <li>A consortium led by <a href="https://www.computerweekly.com/news/366638132/South-Korea-debuts-foundation-model-in-sovereign-AI-push">SK Telecom has built a sovereign AI model</a> designed to reduce reliance on foreign tech, lower costs for local industry and propel South Korea into the top ranks of AI powers.</li> <li>Lenovo’s CIO Playbook 2026 reveals that 96% of APAC organisations are <a href="https://www.computerweekly.com/news/366637305/APAC-enterprises-to-boost-AI-spend-by-15-in-2026">planning to invest more in AI</a>, with a growing reliance on hybrid infrastructure to manage rising inference costs.</li> </ul> </div> </div> </section> Neocloud and sovereign cloud providers offer alternatives to hyperscalers for AI infrastructure and data sovereignty, but availability gaps and a lack of managed AI services can pose challenges to enterprise customers https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/cloud-computing-2-jijomathai-adobe.jpg https://www.computerweekly.com/news/366639689/Weighing-the-trade-offs-of-neoclouds-and-sovereign-clouds Wed, 04 Mar 2026 22:34:00 GMT <p>The physical limits of on-premise storage and compute capacity originally drove the enterprise shift to the cloud, where users have access to “practically infinite” resources, according to Chalan Aras, senior vice-president and general manager of acceleration at Riverbed.</p> <p>Consequently, petabytes of data are now stored in the cloud and organisations are eager to apply <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence (AI)</a> to put this information to work.</p> <p>However, this data may not be in the ideal location for AI processing. Under a <a href="https://www.computerweekly.com/ehandbook/Making-multi-cloud-work">multicloud strategy</a>, enterprise data is inherently spread across multiple providers. Even if all the data required for an AI project resides in a single cloud, it may be stored in a region where power is expensive and therefore unlikely to be equipped with the graphics processing units (GPUs) needed for AI workloads. Either way, enterprises face the daunting task of moving massive quantities of data.</p> <p>Such data movements are expensive, costing up to $80,000 per petabyte in egress fees alone, Aras warned – even when transferring within a single cloud provider. Furthermore, transfers must be strictly governed to ensure the right data reaches the right destination entirely intact. Speed is another major bottleneck – transferring just 1PB over a 10Gbps connection takes around nine days.</p> <p>That’s just the historical data. It is usual to keep feeding the AI model with the most recent data, perhaps once or twice a day. The volumes are much smaller, but it is still important to do the transfers quickly and efficiently as the model is running round the clock, and governance is still essential.</p> <p>Riverbed is now taking its 25 years of data movement experience and applying it to customers’ cloud environments, Aras explained. The process involves extracting data from storage and optimising it for network transfer. “We’re serving it on a plate,” he said.</p> <p>In one instance, an organisation needed to transfer 1PB of data to a new location for AI training but found its existing processes would take 12 days. This was only the first parcel of data, with a further 20PB still to move. The organisation had already booked highly sought-after GPU time and the looming deadline was at risk. Riverbed completed the entire task in three to four weeks, rather than the projected eight to nine months, ensuring data transfer was no longer the project’s limiting factor.</p> <p>Similarly, following a merger in the financial services sector, a company needed to transfer roughly 30PB of data from one cloud to another. Riverbed completed the migration in just over a month while meeting the required governance standards.</p> <p>More broadly, IT teams at established enterprises have historically made data storage and processing decisions based on the circumstances of the time. Layered on top are various decisions made at a departmental or line-of-business level. As a result, businesses today typically operate a mix of on-premise datacentres, multiple cloud environments and numerous <a href="https://www.computerweekly.com/resources/Software-as-a-Service-SaaS">software-as-a-service (SaaS)</a> applications.</p> <p>While consolidating to a single cloud provider is possible, organisations must decide if one provider can truly meet all their needs without forcing compromises. Not even the largest hyperscalers have a presence in every geography, Aras pointed out. For this and other reasons, a second provider is often necessary, even if it means sacrificing the simplicity of a single contract and a single set of skills.</p> <p>Distributing systems across multiple locations is rarely an issue until all that data needs to be aggregated in one place. As AI adoption spreads, this requirement is becoming increasingly common. To extract full value from their data, businesses face a very real need to move large volumes of it – not just as a one-off, but on an ongoing basis.</p> <p>This is especially true for <a href="https://www.computerweekly.com/news/366639619/Emerging-markets-prioritise-top-line-growth-with-agentic-AI">agentic AI</a>, which may need to pull information from a myriad of sources to effectively respond to a prompt, Aras noted. This is great for users, he said, as they can get very quick answers, but it does require the frequent movement of data.</p> <p>Until recently, much of Riverbed’s business revolved around supporting one-time data transfers, such as migrating systems from on-premise locations to the cloud. However, Aras noted that customers increasingly need to move substantial amounts of data continuously to feed their AI strategies. Riverbed’s approach makes its products suitable for both situations, he said.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in APAC</h3> <ul class="default-list"> <li>NetApp’s regional chief discusses the&nbsp;<a href="https://www.computerweekly.com/news/366639344/APAC-firms-still-in-AI-test-mode-as-data-readiness-issues-persist">gap between AI intent and production</a>, the rise of neoclouds, and why the storage firm is counting on getting data AI-ready to win market share.</li> <li>The Australian government has struck a&nbsp;<a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">major five-year volume sourcing agreement with Microsoft</a>&nbsp;to speed up adoption of AI and cloud technologies across the public sector.</li> <li>Singtel and Nvidia have teamed up on a multimillion-dollar facility to&nbsp;<a href="https://www.computerweekly.com/news/366639492/Singtel-Nvidia-to-help-scale-enterprise-AI-deployments">help organisations scale enterprise AI deployments</a>, tackle extreme datacentre power densities, and prepare for the era of embodied AI.</li> <li>A consortium led by&nbsp;<a href="https://www.computerweekly.com/news/366638132/South-Korea-debuts-foundation-model-in-sovereign-AI-push">SK Telecom has built a sovereign AI model</a>&nbsp;designed to reduce reliance on foreign tech, lower costs for local industry and propel South Korea into the top ranks of AI powers.</li> </ul> </div> </div> The cost, speed, and governance of moving petabytes of data across hybrid and multicloud environments is becoming a challenge for enterprises looking to harness the benefits of AI https://cdn.ttgtmedia.com/visuals/German/article/cloud-data-storage-adobe.jpg https://www.computerweekly.com/news/366639742/Scattered-data-cloud-transfers-creating-challenges-in-enterprise-AI Tue, 03 Mar 2026 21:37:00 GMT <p>For all the promises of <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence</a> (AI) automating routine tasks, businesses in developing economies in Southeast Asia are faced with an economic reality – that it’s still cheaper to hire a human than to run an AI agent.</p> <p>This is changing how businesses in the region prioritise agentic AI initiatives, where projects that drive top-line revenue are favoured over those that automate processes to improve productivity and reduce cost.</p> <p>That was according to Arun Kumar Parameswaran, Salesforce’s executive vice-president and managing director for South and Southeast Asia, who highlighted the differences in agentic AI adoption across different markets in the region during a recent interview with Computer Weekly.</p> <p>“Singapore still operates as a mature market for most use cases, where the cost of the agent is far lower than the cost of the human itself,” said Parameswaran.</p> <p>“But as you go into the rest of ASEAN, you have this very interesting dynamic where the cost of doing something with AI is still more than the cost of hiring a human.”</p> <p>Against this backdrop, more enterprises in the region, including Salesforce itself, are deploying AI agents to tackle untouched revenue opportunities, such as sales development agents that score archived leads and book appointments without human intervention.</p> <p>“What we’re seeing in this part of the world is that those top-line use cases seem to get green-lighted faster than productivity, because of the pricing economics in this region,” said Parameswaran.</p> <section class="section main-article-chapter" data-menu-title="Escaping ‘pilot purgatory’"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Escaping ‘pilot purgatory’</h2> <p>The demand for top-line value in emerging markets reflects the general frustration with generative AI. Despite heavy investments in underlying infrastructure, many enterprises only <a href="https://www.techtarget.com/searchcio/post/Why-enterprises-shouldnt-accept-good-enough-AI-ROI">achieve modest AI gains</a>. According to a BCG study, finance market leaders reported AI returns of about 10%, which is barely above the cost of capital.</p> <p>Srini Tallapragada, president and chief engineering and customer success officer at Salesforce, who has spent recent weeks interacting with global CIOs and board members, noted that many organisations are also increasingly tired of running AI pilots.</p> <p>“I call it ‘pilot purgatory’,” said Tallapragada. “Everybody is cool to do quick demos, but if you ask a C-level executive or the board, ‘Did it have an impact on your P&amp;L?’, it’s not there,” he added, referring to profit-and-loss statements that reflect an organisation’s financial performance.</p> <p>The crux of the ROI issue lies in what Tallapragada refers to as the last-mile gap: a deficit in trusted data context, workflow integration, and guardrails required to make AI agents safe and effective in delivering business outcomes.</p> <p>To bridge the gap, Salesforce is urging the industry to rethink how it measures AI consumption. Currently, the industry relies on tokens, the basic units of data processed by a large language model, which Tallapragada argued is a flawed metric for business value.</p> <p>“People are measuring tokens, which are like CPUs – important but not showing value,” he said. Instead, Salesforce is tracking <a href="https://www.constellationr.com/insights/news/salesforces-agentic-work-unit-what-you-need-know">agentic work units</a>, a metric that measures the actual tasks completed by an agent. According to Tallapragada, Salesforce customers have already executed over 2.4 billion agentic work units, which increased by 57% in the fourth quarter alone.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in APAC</h3> <ul class="default-list"> <li>NetApp’s regional chief discusses the <a href="https://www.computerweekly.com/news/366639344/APAC-firms-still-in-AI-test-mode-as-data-readiness-issues-persist">gap between AI intent and production</a>, the rise of neoclouds, and why the storage firm is counting on getting data AI-ready to win market share.</li> <li>The Australian government has struck a <a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">major five-year volume sourcing agreement with Microsoft</a> to speed up adoption of AI and cloud technologies across the public sector.</li> <li>Singtel and Nvidia have teamed up on a multimillion-dollar facility to <a href="https://www.computerweekly.com/news/366639492/Singtel-Nvidia-to-help-scale-enterprise-AI-deployments">help organisations scale enterprise AI deployments</a>, tackle extreme datacentre power densities, and prepare for the era of embodied AI.</li> <li>A consortium led by <a href="https://www.computerweekly.com/news/366638132/South-Korea-debuts-foundation-model-in-sovereign-AI-push">SK Telecom has built a sovereign AI model</a> designed to reduce reliance on foreign tech, lower costs for local industry and propel South Korea into the top ranks of AI powers.</li> </ul> </div> </div> <p>As more AI projects move from pilots to production, the software industry is also grappling with how to charge for autonomous agents. Software-as-a-service (SaaS) suppliers are currently caught between offering consumption-based models and traditional, predictable licensing.</p> <p>“The pendulum has swung quite dramatically back and forth,” said Parameswaran. “Consumption-based pricing is highly non-predictable. If you remember the early days of the cloud where people had runaway bills, we’re starting to see elements of that in the AI world. Customers want the best of both worlds: they want consumption-based, but they also want it to be predictive.”</p> <p>To accommodate this, Salesforce is currently offering a mix of pay-as-you-go consumption models and an agentic enterprise licence agreement (AELA) for customers that want unlimited use of consumption-based services, such as Agentforce, Slack and MuleSoft, for a fixed fee over two or three years.</p> <p>Lisa Singer, principal analyst and vice-president of Forrester, a market research firm, noted that licensing models like AELA show that the value of AI agents is not correlated to usage volume, but to the economic outcomes they enable.</p> <p>“Accepting short‑term unprofitability only makes sense when a vendor believes that agents materially reshape enterprise cost structures, productivity or growth,” she <a target="_blank" href="https://www.forrester.com/blogs/ai-agents-become-economic-actors-salesforce-rewrites-the-rules-of-pricing/" rel="noopener">wrote in a blog post</a>. “The pricing signals confidence that agent-driven value is durable and monetisable over time.”</p> </section> <section class="section main-article-chapter" data-menu-title="The real bottleneck"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The real bottleneck</h2> <p>As multi-agent ecosystems mature, enterprises will eventually need a so-called agent fabric or control plane to govern interactions between AI agents built by different suppliers, such as a Salesforce customer service agent needing to query an <a href="https://www.computerweekly.com/news/366632305/SAP-AI-is-commoditising-business-applications">SAP supply chain agent</a> via the <a href="https://www.computerweekly.com/news/366622317/Google-offers-open-protocol-for-AI-agent-connectivity">agent-to-agent protocol</a>.</p> <p>“Everybody is great on day one. I always say the real work starts on day two, day three, day 100,” said Tallapragada. “These agents are running, but who is ensuring they’re running properly? Just like humans, they need to be managed. They can drift. You need traceability, auditability and observability.”</p> <p>Crucially, as these governance and economic models take shape, the barriers to agentic AI adoption are changing as well. While compute capacity once hindered adoption, the challenge today lies in the enterprises themselves.</p> <p>“If you asked me a year back, the biggest issue was getting GPUs [graphics processing units]. That is gone, frankly,” said Tallapragada. “The bigger constraint is customers trying to identify use cases where they generate value instead of demos. It’s the internal change management and <a href="https://www.computerweekly.com/news/366593419/IBM-Reimagine-processes-to-unlock-AIs-value">business process reengineering</a> – that’s the bottleneck more than infrastructure.”</p> </section> While firms in mature markets are using AI agents to automate routine tasks, those in emerging markets where the cost of the technology is higher than that of human labour are favouring revenue-generating use cases https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/Bangkok-Thailand-night-market-Stockbym-adobe.jpg https://www.computerweekly.com/news/366639619/Emerging-markets-prioritise-top-line-growth-with-agentic-AI Tue, 03 Mar 2026 06:15:00 GMT <p>Brian Stafford, CEO of governance, risk and compliance specialist Diligent, has been talking to his clients about <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence (AI)</a> for years. Two years ago, it was still a relatively new concept for boards and the C-suite, he recalled. One year ago, some were still holding back from the technology. Now, he said, it is no longer about the “what”, but the “how”.</p> <p>Organisations are actively asking how they can use AI to drive innovation, efficiency and effectiveness. According to Stafford, sometimes it is the board that pushes this agenda while other times, it is the CEO.</p> <p>He gave the example of the CEO of a very large company who tried to <a href="https://www.computerweekly.com/news/366625258/AI-should-be-used-to-improve-boardroom-efficiency">get his board to use all of Diligent’s AI capabilities</a>, only to be told they were happy as they were. Ironically, that same CEO was being pushed by the board to deliver corporate AI transformation, even though the board was unwilling to undergo its own. Ultimately, transformation has to start “at the top of the house”, Stafford argued, whether that means the CEO, or the CEO and the board together.</p> <p>Recent Diligent research into governance leaders in Australia showed that only 43% ranked AI as a strategic priority. Stafford wondered how the remaining 57% could fail to consider it a priority but conceded that “the data is the data”.</p> <p>“It needs to be [a priority], and that’s the way we pushed and steered our road map for our clients,” he said. “If you asked CEOs, the number would probably be higher than that.”</p> <p>Globally, Stafford finds adoption and interest in AI among boards to be nearly as high in the UK, continental Europe and Australia as it is in the US. He attributed this partly to <a href="https://www.computerweekly.com/feature/ChatGPT-Everything-you-need-to-know">ChatGPT becoming the fastest-growing consumer tool</a> in history, which led to widespread awareness of AI’s potential.</p> <p>“The organisations that are further ahead are the ones where the CEOs are leading the change,” he said. “I see those CEOs here in Australia, in the US, in the UK, in continental Europe. I have not seen a radical change or difference across geography.”</p> <p>Meanwhile, corporate obligations continue to grow in areas such as compliance, legal, audit and risk management. As Stafford pointed out, for most companies, it’s not a badge of honour to say their compliance spend is growing faster than their revenue. To become more efficient in meeting those obligations when there is always more work to do, the only answer is to use AI to help drive automation. Many of Diligent’s clients say using the firm’s AI products allows them to become more strategic.</p> <p>The Diligent report also revealed that 13% of boards have recruited directors with AI expertise. Asked whether this figure is surprisingly high, Stafford pointed out that a number of clients have picked up directors with experience from Silicon Valley and other tech hubs to play an early role in AI transformation.</p> <p>However, many boards struggle with the trade-off of adding an individual who may have deep AI expertise when most board members traditionally possess more rounded profiles. Stafford noted that there are not many people who combine deep technical expertise in AI with a broader business perspective. Consequently, existing board members need to learn new ways of interacting with their peers when younger people, or those with specific expertise in a new field, join their ranks.</p> <p>Furthermore, the report found that only 21% of respondents’ organisations mandate AI training for directors. Stafford identified three aspects of AI that directors need to understand, starting with grasping the use and <a href="https://www.techtarget.com/searchitoperations/opinion/Essential-transformations-to-escape-AI-stagnation">power of AI for transforming internal operations</a> and delivering products and services.</p> <p>Second, they need to get to grips with <a href="https://www.techtarget.com/searchenterpriseai/definition/AI-governance">AI governance</a>, which involves navigating the complexity of the technology and creating the right structures, guardrails and controls. Finally, directors must consider <a href="https://www.computerweekly.com/feature/How-AI-ethics-is-coming-to-the-fore-with-generative-AI">AI ethics</a>, focusing on the long-term impact and obligations the technology will bring.</p> <p>Stafford noted that while training sometimes glances over those three areas, it can also end up focusing deeply on just one specific element. Ultimately, board members must ensure they are aware of where and how AI is being applied in their organisation. They need to confirm that an appropriate data framework is in place and that AI projects are operating strictly within it.</p> <p>Another important job for the board is ensuring the right controls and governance frameworks surround the use of AI, though Stafford acknowledged this remains problematic in the absence of generally agreed or mandated industry standards.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in Australia</h3> <ul class="default-list"> <li>The DTA has struck a major five-year <a href="https://www.computerweekly.com/news/366639595/Australia-inks-five-year-deal-with-Microsoft-to-drive-AI-and-cloud-adoption">volume sourcing agreement with Microsoft</a> to speed up adoption of&nbsp;AI and cloud&nbsp;technologies across the public sector.</li> <li>Wesfarmers has signed a&nbsp;<a href="https://www.computerweekly.com/news/366639137/Wesfarmers-to-deploy-agentic-AI-in-retail-operations">multi-year deal with Google Cloud</a>&nbsp;to deploy&nbsp;agentic AI&nbsp;across its portfolio of brands, including Kmart, Officeworks, Priceline and OnePass.</li> <li>ANZ to be the&nbsp;<a href="https://www.computerweekly.com/news/366638802/ANZ-rolls-out-AI-agents-for-business-bankers">first in Asia-Pacific to deploy Salesforce Agentforce</a>&nbsp;at scale, following a national roll-out of a CRM platform that consolidates data from different systems to ease administrative toil.</li> <li>US chip startup Groq plans to invest up to $300m to provide Australian businesses with&nbsp;<a href="https://www.computerweekly.com/news/366634613/Groq-commits-up-to-300m-in-Australia-expansion">compute capacity for AI inferencing</a>&nbsp;and help solve issues around data sovereignty for major users like Quantium.</li> </ul> </div> </div> While C-suite interest in AI has shifted from the ‘what’ to the ‘how’, Diligent CEO Brian Stafford warns that true enterprise transformation requires hands-on leadership from the board https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/meeting-boardroom-business-adobe.jpg https://www.computerweekly.com/news/366639510/AI-transformation-must-start-at-the-top-but-boards-remain-divided Mon, 02 Mar 2026 19:28:00 GMT <p>Micron Technology has opened its first semiconductor test and assembly facility in Sanand, Gujarat, representing a combined investment of about $2.75bn by the chipmaker and its government partners.</p> <p>The facility converts advanced DRAM and NAND wafers from Micron’s global manufacturing network into finished memory and storage products. Once fully ramped up, the first phase of the Sanand operation will feature more than 500,000 ft² of cleanroom space, making it one of the world’s largest single-floor test and assembly cleanrooms.</p> <p>The ISO-certified site has already begun commercial production. To mark the opening, Micron presented its first shipment of made-in-India memory modules to Dell Technologies for use in locally manufactured laptops. Micron expects to test and assemble tens of millions of chips at the Sanand plant in 2026, scaling to hundreds of millions in 2027.</p> <p>Sanjay Mehrotra, president and CEO of Micron Technology, described the launch as a “proud moment” for the company and India’s growing semiconductor industry. “This pioneering facility, the first assembly and test site of its kind in the country, helps build a resilient ecosystem that underpins the global AI [artificial intelligence] economy,” he said.</p> <p>Speaking at the facility’s inauguration, which was <a href="https://www.youtube.com/watch?v=UslbyeIoKaA">broadcast live to the public on YouTube</a>, Indian prime minister Narendra Modi noted that the opening was a major leap forward for the country’s technological ambitions.</p> <p>“While the <a href="https://www.computerweekly.com/news/366639543/India-AI-Impact-Summit-Open-source-gains-ground-but-sovereignty-tensions-persist">AI Summit</a> introduced the world to India’s AI prowess, today is a testament to India’s commitment to technology leadership,” he said. “India, long known for its software strength, is now firmly establishing its identity in the hardware sector as well. If oil was the regulator of the last century, microchips will be the regulator of this century.”</p> <p>Addressing the global technology sector directly, Modi added: “India has just one message for investors around the world: India is ready, India is reliable, India delivers.”</p> <p>Ashwini Vaishnaw, India’s union minister for railways, communications, electronics and IT, noted that the inauguration marks a major step for the country’s chip production capabilities. “India is now moving from being a consumer of chips to becoming a global hub for semiconductor manufacturing and innovation,” he said.</p> <section class="section main-article-chapter" data-menu-title="Global supply chain resilience"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Global supply chain resilience</h2> <p>The expansion of test and assembly operations in India is part of broader efforts by Micron to diversify its global manufacturing footprint and meet the growing market demand for storage memory driven by the boom in AI and data-centric applications.</p> <p>In January 2026, Micron started constructing a <a href="https://www.computerweekly.com/news/366637651/Micron-breaks-ground-on-24bn-Singapore-wafer-fab">$24bn advanced wafer fabrication facility in Singapore</a> that will produce NAND flash memory. The Singapore expansion will feature the country’s first double-storey wafer fab and provide an additional 700,000ft² of cleanroom space when it becomes operational in 2028.</p> <p>Just weeks prior, Micron also formally broke ground on a <a href="https://www.constructiondive.com/news/micron-break-ground-new-york-semiconductor-fab/809171/">new megafab in Clay, New York</a>. While the Singapore facility focuses on NAND flash chips, the New York complex will produce DRAM system memory. Micron expects to invest up to $100bn over the next two decades to build what it claims will be the largest semiconductor facility in the US.</p> <p>By increasing capacity across North America, Southeast Asia, and now India, the company aims to build a more resilient supply chain capable of weathering geopolitical friction and regional disruptions.</p> <p>To support its Gujarat operations, Micron is partnering with Indian institutions, including Pandit Deendayal Energy University and Namtech, as well as government-sponsored skills development programmes. These initiatives are focused on science, technology, engineering and mathematics education, specialised training for advanced manufacturing roles, and digital AI literacy across the region.</p> <p>In line with the chipmaker’s environmental commitments, the Sanand facility has been designed to meet Leadership in Energy and Environmental Design gold standards, and uses advanced water-saving technologies to achieve zero liquid discharge.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about IT in India</h3> <ul class="default-list"> <li>Nvidia unveils <a href="https://www.computerweekly.com/news/366639266/Nvidia-backs-Indias-sovereign-AI-push-with-gigawatt-scale-infrastructure">compute expansion with L&amp;T, Yotta and E2E Networks</a> at the India AI Impact Summit, paving the way for domestic heavyweights to build AI agents and physical AI applications.</li> <li>While <a href="https://www.computerweekly.com/news/366639543/India-AI-Impact-Summit-Open-source-gains-ground-but-sovereignty-tensions-persist'">open source AI gained unprecedented recognition</a> during the India AI summit, divisions over governance, market concentration and regulatory power cast doubt on whether the technology will benefit society as a whole.</li> <li>Krutrim, the AI venture from Indian ride-hailing and EV giant Ola, is building a&nbsp;<a href="https://www.computerweekly.com/news/366629172/Olas-Krutrim-builds-AI-first-sovereign-cloud-for-India">vertically integrated sovereign cloud and AI stack for India</a> using Cloudera’s data platform as a core component of its architecture.</li> <li>OpenAI is preparing a&nbsp;<a href="https://www.computerweekly.com/news/366630088/OpenAI-targets-India-with-datacentre-push">one-gigawatt datacentre in India</a>&nbsp;as part of its $500bn Stargate infrastructure push, in what would be its biggest bet yet on its second-largest user base.</li> </ul> </div> </div> </section> Test and assembly site in Gujarat marks India’s first commercial semiconductor production following multibillion-dollar investments in US and Singapore to meet growing demand for storage and memory chips https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/semiconductor-silicon-wafer-chip-IM-Imagery-adobe.jpg https://www.computerweekly.com/news/366639653/Micron-opens-275bn-chip-assembly-plant-in-India Mon, 02 Mar 2026 04:55:00 GMT <p>Even as tech suppliers like OpenAI and Nvidia are pouring billions into <a href="https://www.computerweekly.com/feature/Top-AI-infrastructure-considerations">artificial intelligence (AI) infrastructure</a>, much of the corporate sector has not moved past the experimental stage on the AI adoption curve, according to Andrew Sotiropoulos, senior vice-president and general manager for Asia-Pacific at NetApp.</p> <p>Speaking to Computer Weekly on the sidelines of a customer event in Singapore, Sotiropoulos said that while there are many companies that are testing the use of AI in specific projects, very few are putting “pedal to the metal” and going into full production.</p> <p>The bottleneck, he noted, isn’t a lack of ambition, but the state of enterprise data. For AI models to be effective, they require vast amounts of clean, organised information, but many companies are sitting on fragmented data silos that are difficult to process.</p> <p>“That is the single biggest issue everyone has, even internally. It is all about cleansing the data and getting it AI-ready,” he said, calling the ability to solve this problem the industry’s “ticket to the dance”.</p> <p>To address this, NetApp has rolled out a software suite called <a href="https://www.techtarget.com/searchstorage/opinion/NetApp-aims-to-transcend-storage-roots-with-vision-to-make-data-AI-ready">AI Data Engine</a> (AIDE) that provides semantic search, data vectorisation and data guardrails to simplify the entire AI data pipeline by collapsing multiple data preparation and management steps.</p> <p>Through data change detection and data synchronisation, AIDE also helps to eliminate redundant copies and keeps data up to date. More importantly, it allows data scientists to focus on model training rather than manual data preparation while ensuring that the data feeding the models is clean and traceable, a requirement for enterprise compliance.</p> <p>Meanwhile, NetApp has also introduced <a href="https://www.techtarget.com/searchstorage/opinion/NetApp-aims-to-transcend-storage-roots-with-vision-to-make-data-AI-ready">AFX</a>, a disaggregated storage architecture that lets enterprises scale their storage and compute independently. This helps improve the economics of AI workloads, preventing organisations from overprovisioning storage compute and improving resource efficiency.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about storage in APAC</h3> <ul class="default-list"> <li>Vast Data CEO Renen Hallak explains how the company’s unified data platform that <a href="https://www.computerweekly.com/news/366573814/How-Vast-Data-is-simplifying-data-infrastructure">combines storage, database and compute capabilities into a single technology stack</a> can improve the efficiency of analytics and AI workloads.</li> <li><a href="https://www.computerweekly.com/feature/How-open-source-is-shaping-data-storage-management">Open source data storage</a> offers a great deal of flexibility, but unlocking its benefits will require strong technical resources to meet requirements such as stability, high availability and security.</li> <li>Pure Storage’s head of AI infrastructure, Par Botes, argues that the <a href="https://www.computerweekly.com/news/366632030/How-AI-is-driving-a-rethink-of-storage-architecture">growing use of AI will require storage systems with audit trails</a> and versioning to ensure trust and traceability.</li> <li><a href="https://www.computerweekly.com/feature/What-are-the-storage-requirements-for-AI-training-and-inference">Storage for AI</a> must cope with huge volumes of data that can multiply rapidly as vector data is created, plus lightning-fast I/O requirements and the needs of agentic AI.</li> </ul> </div> </div> <p>For now, AIDE runs with AFX on-premise, but NetApp aims to expand their reach to public clouds and other locations, including <a href="https://www.computerweekly.com/feature/Neoclouds-Meeting-demand-for-AI-acceleration">neoclouds</a> from sovereign cloud providers such as Singapore’s <a href="https://www.bitdeer.com/">Bitdeer</a> and India’s <a href="https://yotta.com/">Yotta</a>.</p> <p>“The sovereign cloud play is the next wave,” said Sotiropoulos, noting that neocloud providers are gaining traction by addressing data residency and sovereignty requirements that global players sometimes struggle with. Neocloud providers are also building up their AI capabilities as a Trojan horse to build relationships with customers, he added.</p> <p>The growing demand for AI infrastructure, however, is placing pressure on NetApp’s channel partners. Sotiropoulos admitted that traditional distributors and systems integrators face a “knowledge, speed and capability” gap. He urged partners to secure a “seat at the table” by upskilling their workforce to manage the full AI solution stack, rather than just the hardware components.</p> <p>NetApp claims to be gaining market share. Citing recent <a href="https://my.idc.com/getdoc.jsp?containerId=prUS54034425">market data from analyst firm IDC</a>, Sotiropoulos said the company has narrowed the gap with its competitors in both all-flash and total enterprise storage segments. Globally, NetApp ranked third in terms of revenue during the third quarter of 2025, behind Dell Technologies and Huawei.</p> <p>Sotiropoulos attributed NetApp’s growth to its unified data strategy, which allows customers to replicate data management policies across on-premise and public cloud environments, along with AI operations and cyber resiliency capabilities.</p> <p>His comments came as NetApp brought its flagship customer event to Singapore for the first time, a move that reflects the city-state’s position as a gateway to the broader ASEAN market.</p> <p>While the company previously flew customers to its global event in Las Vegas for such updates, budget constraints and the difficulty faced by customers in securing travel approvals have shifted its engagement strategy towards regional hubs.</p> <p>“Our vision is that ASEAN as a hub is a great opportunity to bring customers from countries like Thailand and Indonesia into Singapore,” he said. “It wasn’t because we have a unique strategy just for ASEAN now, but it was the right time for us to focus on this particular region using Singapore as a hub.”</p> NetApp’s regional chief discusses the gap between AI intent and production, the rise of neoclouds, and why the storage firm is counting on getting data AI-ready to win market share https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/Storage-abstract-data-2-adobe.jpg https://www.computerweekly.com/news/366639344/APAC-firms-still-in-AI-test-mode-as-data-readiness-issues-persist Mon, 02 Mar 2026 01:27:00 GMT <p>When I joined <a href="https://www.computerweekly.com/news/366631832/Alibaba-Cloud-targets-full-stack-AI-dominance">Alibaba</a> in 2017, one of the first things that surprised me was how little my Western colleagues understood about what was happening in China’s technology landscape. The prevailing narrative in boardrooms from London to New York was simple – China copies, the West innovates. That assumption was already wrong then. Today, it is dangerously outdated.</p> <p>As <a href="https://www.computerweekly.com/news/366637505/JPMorgan-CEO-urges-slowdown-of-AI-roll-out-to-save-society">Western enterprises race to integrate generative artificial intelligence (AI)</a> into their operations, many continue to view China’s AI ambitions through a competitive lens shaped by outdated stereotypes. This blind spot is not just an intellectual failing – it is a strategic one, and it is costing companies real opportunities and leaving them vulnerable to disruptions they never saw coming.</p> <section class="section main-article-chapter" data-menu-title="It was never about catching up"> <h2 class="section-title"><i class="icon" data-icon="1"></i>It was never about catching up</h2> <p>The most persistent misunderstanding is that China’s AI strategy is about replicating what Silicon Valley has built.</p> <p>In reality, China’s approach to AI is architecturally different. Where Western tech companies have largely pursued AI as a product category – chatbots, copilots, and standalone tools that can be sold to enterprises – China has treated AI as infrastructure: a utility layer woven into the fabric of commerce, logistics, government services, and daily life.</p> <p>During my time at Alibaba, I watched generative AI being deployed not as a flashy demo, but as an invisible engine powering everything from real-time product recommendations to automated merchant communications at a scale most Western companies have never attempted.</p> <p>By 2019, AI-generated product descriptions were already standard practice across the platform. The West did not reach a comparable level of mainstream AI integration in commerce until the <a href="https://www.computerweekly.com/opinion/All-I-want-for-Christmas-is-a-ChatGPT-nativity-scene">ChatGPT era</a>, roughly four years later.</p> <p>This distinction matters. Chinese companies were not waiting for a perfect model before deploying. They were building deployment-first cultures where AI was tested in live, high-stakes environments from day one.</p> <p>The Western approach of perfecting the model in the lab before cautiously rolling it out is, to many Chinese AI leaders, an unnecessarily slow way to learn.</p> </section> <section class="section main-article-chapter" data-menu-title="Second mover’s advantage"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Second mover’s advantage</h2> <p>We have all heard of the “first mover’s advantage”, where if you are the first to market, it is easier to establish an affinity for your brand with the customer. But sometimes, it might be the second mover that wins. China follows this strategy that I call second mover’s advantage. Take what has been developed and released and optimise on top of it. China basically did it again in the AI space – the West saw this in action <a href="https://www.sharongai.com/post/deepseek-searching-for-answers-in-the-depth-of-the-us-china-ai-war">when DeepSeek came out of China and shook the stock markets</a>, sending Nvidia’s share price tumbling.</p> <p>From the manufacturing of goods to algorithms, no matter if it’s hardware or software, it seems that China is remarkably good at taking something that might have emerged from the West, optimising it, and making it cheaper and more accessible to the masses. Is it better ultimately? That one’s debatable.</p> </section> <section class="section main-article-chapter" data-menu-title="The ecosystem, not the algorithm"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The ecosystem, not the algorithm</h2> <p>Western companies tend to fixate on who has the best foundational model. The assumption is that the <a href="https://www.computerweekly.com/opinion/Forget-AGI-business-leaders-are-still-trying-to-figure-out-how-to-make-AI-work">race for artificial general intelligence</a> will be won by whichever company produces the most capable large language model. China’s strategy suggests a fundamentally different theory of victory.</p> <blockquote> <div class="imagecaption alignLeft"> <img src=" https://cdn.ttgtmedia.com/rms/computerweekly/Sharon-Gai-140x180.jpg" alt="Photo of author Sharon Gai"> </div> <p><span style="font-size: 14pt;"><strong><span style="color: #34495e;">“Western companies that continue to misread China’s approach are not just underestimating a competitor. They are misunderstanding the game itself”</span></strong></span></p> <p><em><span style="color: #34495e;">Sharon Gai</span></em></p> </blockquote> <p>Rather than concentrating resources on a single frontier model, China has invested in building an ecosystem where AI capabilities are distributed across industries. The government’s AI development plans, updated repeatedly since 2017, are explicit about this – the goal is not to produce one dominant AI company, but to create an environment where thousands of companies can apply AI to specific problems in manufacturing, agriculture, healthcare, and urban management.</p> <p>For Western IT leaders, this has practical implications. When evaluating Chinese competitors or potential partners, looking only at their model benchmarks misses the point. The competitive advantage often lies in the integration layer – in how tightly AI is embedded into supply chains, customer journeys, and operational workflows. A Chinese company with a seemingly modest model but deep application integration will often outperform a Western competitor with a more powerful model that sits in a silo.</p> </section> <section class="section main-article-chapter" data-menu-title="Data as a national resource"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Data as a national resource</h2> <p>Another common blind spot is the <a href="https://www.computerweekly.com/news/366637344/AI-governance-provides-guardrails-for-faster-innovation">role of data governance</a>. Western companies often frame data privacy and AI capability as a binary trade-off – either you protect privacy or you build powerful AI. China’s approach complicates this framing.</p> <p>While the country’s data regulations are different from the General Data Protection Regulation (GDPR), they are neither absent nor unsophisticated. China’s Personal Information Protection Law (PIPL) introduced significant data protections in 2021, and enforcement has been meaningful.</p> <p>What is different is how data flows between the public and private sectors. In China, data is treated more like a shared national resource than a proprietary corporate asset. This creates feedback loops that simply do not exist in Western markets, where data silos between companies, government, and institutions are often reinforced by regulation and competitive incentives.</p> </section> <section class="section main-article-chapter" data-menu-title="What this means for Western leaders"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What this means for Western leaders</h2> <p>The practical takeaway is not that Western companies should replicate China’s model. The regulatory, cultural, and political contexts are too different for direct imitation. But there are strategic lessons worth absorbing.</p> <p>First, stop treating AI deployment as something that happens after the model is “ready”. The companies gaining the most ground, in China and increasingly elsewhere, are the ones deploying imperfect AI in controlled but real environments, learning from live data, and iterating rapidly. Waiting for perfection is a luxury that the pace of competition no longer affords.</p> <p>Second, invest in the integration layer. The model is only as valuable as the ecosystem it connects to. Western organisations that focus exclusively on procuring the best model while neglecting the workflows, data pipelines, and cultural changes needed to make that model useful will find themselves outpaced by competitors who build tighter loops between AI and operations.</p> <p>Third, develop genuine China literacy within your strategy teams. Too many Western companies rely on surface-level reporting or outdated assumptions about Chinese technology. The executives who will navigate the next decade of AI competition successfully are those who invest in understanding what is actually happening on the ground, not what fits the familiar narrative.</p> <p>The AI race is not a single sprint with one finish line. It is a complex, multi-front contest where different strategies can win in different domains. Western companies that continue to misread China’s approach are not just underestimating a competitor. They are misunderstanding the game itself.</p> <hr> <p><a href="https://www.sharongai.com/"><i>Sharon Gai</i></a><i> is an AI transformation strategist, keynote speaker, and former Alibaba executive. She is the author of “</i><a href="https://www.amazon.com/How-More-Less-Future-Proofing-AI-driven/dp/1394352360"><i>How to do more with less using AI</i></a>”<i> and advises Fortune 500 companies on AI adoption and organisational change.</i></p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about the global AI race</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366609796/China-dominates-AI-and-advanced-analytics-research">China dominates AI and advanced analytics research</a>: Research looking at each country’s AI contribution, based on the quantity and quality of researchers papers published, shows China is leading the way.</li> <li><a href="https://www.computerweekly.com/news/366637503/Davos-2026-Smart-thinking-needed-for-sovereign-AI-investment">Davos 2026: Smart thinking needed for sovereign AI investment</a>: Policy-makers need to figure out how they can carve a niche in a world dominated by China and the US.</li> <li><a href="https://www.computerweekly.com/news/366628632/DeepSeek-shows-enterprises-model-distillation-opportunity">DeepSeek shows enterprises model distillation opportunity</a>: DeepSeek showed how it is possible to run an AI model using far less compute than existing models. AI model distillation is now becoming mainstream.</li> </ul> </div> </div> </section> Executives in the West consistently underestimate technological progress in China – and the country’s differing approach to AI development will lead to a significant advantage if Western leaders fail to learn https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/China-ForbiddenCity-adobe.jpg https://www.computerweekly.com/opinion/What-Western-companies-misunderstand-about-Chinas-AI-strategy Fri, 27 Feb 2026 12:25:00 GMT <p>The role of the chief information security officer (CISO) is becoming untenable under the weight of mounting vulnerabilities, complex threats and chronic alert fatigue, according to CrowdStrike president Michael Sentonas.</p> <p>In response, the company is doubling down on <a href="https://www.computerweekly.com/opinion/Generative-and-agentic-AI-in-security-What-CISOs-need-to-know">agentic security</a>, leveraging artificial intelligence (AI) to fundamentally change how <a href="https://www.computerweekly.com/news/252478461/Inside-the-SOC-the-nerve-centre-of-security-operations">security operations centres</a> (SOCs) operate.</p> <p>“We’re at the point where it’s impossible to be effective when you think about all of the complexities that an organisation needs to deal with,” Sentonas told Computer Weekly in a recent interview.&nbsp; “It’s statistically impossible to keep up with all of the vulnerabilities and the patching that happens. And it’s impossible to keep up with the number of threats and threat actors.”</p> <p>To combat this, CrowdStrike has baked in AI capabilities across its entire platform to process telemetry and signals automatically. For its efforts, the company received an award last year from SE Labs, which assesses security products, for achieving 100% detection and prevention of cyber threats, with no false positives in stopping ransomware.</p> <p>In transitioning organisations towards the agentic SOC, CrowdStrike has released a suite of purpose-built AI agents, including threat hunting agents, malware analysis agents and triage agents. It also provides translation agents to help customers migrate from legacy systems, such as Splunk, to CrowdStrike’s next-generation <a href="https://www.computerweekly.com/feature/AI-powered-cloud-SIEM-Real-time-threat-intel-boosts-defences">security information and event management</a> (SIEM) product, Falcon LogScale.</p> <p>Beyond out-of-the-box offerings, CrowdStrike is also empowering security teams to build their own custom automations. For example, through its Charlotte AI AgentWorks no-code platform, customers can use natural language to build specialised agents tailored to their unique environments.</p> <p>According to Sentonas, customers are already running hundreds of thousands of custom automations weekly, building agents for patching, reporting, compliance checks and IT management. “If you can think it and you can script it, you can automate it on the platform,” he said.</p> <section class="section main-article-chapter" data-menu-title="Reshaping the SOC, not replacing it"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Reshaping the SOC, not replacing it</h2> <p>The rise of security agents has inevitably sparked debate about the future of entry-level SOC roles. However, Sentonas noted that agentic security is not about reducing headcount. Instead, it’s about shifting human effort to the “tip of the spear” – the 1-2% of complex attacks that require human intuition and intervention.</p> <p>“When people talk about the agentic SOC, they think they’re not going to need all the people – that’s not true,” he said. “What we’re going to do is improve the quality of life for people in the SOC, because systems are really good at repetitive tasks. What they’re not good at is when adversaries are changing tactics and rolling out complex attacks that AI models may not be good enough today to deal with.”</p> <p>Sentonas warned against the idea of eliminating tier-one security analysts entirely, as that could create a talent pipeline issue for the cyber security industry in the longer term.</p> <p>“If you don’t have level ones, how do you get to level three and four?” he said. “When you’re a level three or level four analyst, you have many years of experience and scars on the back. The technology today is not at a point where we need fewer people; the technology is going to enable people to do more high-value tasks.”</p> <p>While the appetite for AI in cyber security is high, Sentonas acknowledged that organisations are rightfully cautious about the risks of granting AI agents access to sensitive environments. “A big concern that people have with a lot of AI models is a fear of not knowing what they’re doing and not knowing what changes they make,” he noted.</p> <p>To mitigate this, CrowdStrike operates its AI tools under a model of bounded autonomy, enabling security teams to have full oversight of AI-driven decisions, and define when and how automated actions occur, making the technology suited for regulated industries such as banking and finance.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about cyber security in APAC</h3> <ul class="default-list"> <li>Singapore mobilised over 100 cyber defenders to neutralise a sophisticated APT actor which infiltrated Singtel, StarHub, M1 and Simba networks in the country’s&nbsp;<a href="https://www.computerweekly.com/news/366638973/Singapore-mounts-largest-ever-cyber-operation-to-oust-APT-actor">largest coordinated cyber incident response to date</a>.</li> <li>Japan’s Nikkei has confirmed a major data breach that potentially&nbsp;<a href="https://www.computerweekly.com/news/366634243/Nikkei-data-breach-exposes-personal-data-of-over-17000-staff">exposed the personal information of more than 17,000 employees</a>&nbsp;and business partners after hackers infiltrated its internal Slack messaging platform.</li> <li>Australian privacy commissioner warns that the&nbsp;<a href="https://www.computerweekly.com/news/366633983/Fewer-data-breaches-in-Australia-but-human-error-now-a-bigger-threat">human factor is a growing threat</a>&nbsp;as notifications caused by staff mistakes rose significantly even as total breaches declined 10% from a record high.</li> <li>Philippine bank&nbsp;<a href="https://www.computerweekly.com/news/366633428/BDO-Unibank-taps-Zscaler-to-secure-cloud-migration">BDO is shoring up its cyber security capabilities</a>&nbsp;to protect its data and systems as it moves more services to the cloud and expands its physical presence into remote areas of the archipelago.</li> </ul> </div> </div> <p>Ultimately, the effectiveness of AI agents hinges on the data feeding them. Leveraging a decade’s worth of security telemetry, CrowdStrike is also applying its agentic capabilities to threat intelligence.</p> <p>Late last year, it introduced Threat AI, an agentic threat intelligence system designed to automate complex workflows, such as reverse-engineering malware, identifying code similarities and generating <a href="https://www.techtarget.com/searchsecurity/news/252527588/Googles-new-YARA-rules-fight-malicious-Cobalt-Strike-use">Yara pattern-matching rules</a>, at the speed required to counter fast-moving adversaries.</p> <p>“The technology is here; it’s not going away,” said Sentonas. “People that will benefit from the technology in the long term are the ones who lean in today.”</p> <p>According to Gartner, 70% of large SOCs are expected to pilot AI agents to augment their security operations by 2028, but only 15% will achieve measurable improvements without structured evaluations.</p> <p>“The potential of AI agents to transform security operations and ease workloads is real, but only if approached with rigour and evaluated through an outcome-driven lens,” Craig Lawson, vice-president analyst at Gartner, wrote in a <a href="https://www.computerweekly.com/opinion/How-AI-agents-are-driving-the-future-of-security-operations">recent commentary for Computer Weekly</a>.</p> <p>“Every investment should be tied to measurable outcomes such as improvements in <a href="https://www.techtarget.com/searchsecurity/tip/The-best-incident-response-metrics-and-how-to-use-them">mean time to respond</a>&nbsp;and <a href="https://www.techtarget.com/searchsecurity/tip/The-best-incident-response-metrics-and-how-to-use-them">mean time to contain</a>, reduction in false positives or analyst workload,” he said. “Ask vendors for evidence of operational improvements in environments like your own before making any commitments.”</p> </section> By embedding AI agents across its platform, CrowdStrike is looking to help security teams automate repetitive security tasks, enabling them to focus on complex and stealthier threats that could slip under the radar https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/IT-security-fotolia.jpg https://www.computerweekly.com/news/366639370/CrowdStrike-touts-agentic-SOC-to-tackle-security-woes Thu, 26 Feb 2026 02:18:00 GMT <p>Singtel has teamed up with Nvidia to launch an <a href="https://www.computerweekly.com/resources/Artificial-intelligence-automation-and-robotics">artificial intelligence</a> (AI) centre of excellence (CoE) to help organisations overcome the infrastructure and skills bottlenecks that stand in the way of scaling up AI deployments.</p> <p>Announced today, the multimillion-dollar facility is expected to provide a deployment pathway for organisations struggling to move their AI initiatives <a href="https://www.computerweekly.com/news/366627053/Enterprise-AI-adoption-moving-beyond-experimentation">beyond the experimental phase into full-scale production</a>.</p> <p>Bill Chang, CEO of Singtel Digital InfraCo, the telco’s digital infrastructure business, said that unlike other AI CoEs in Singapore, Singtel’s CoE is focused on applied AI, where enterprises bring real-world problem statements and collaborate with an ecosystem of <a href="https://www.computerweekly.com/feature/LLMs-explained-A-developers-guide-to-getting-started">large language model</a>&nbsp;(LLM) makers, application providers, and systems integrators.</p> <p>To help customers scale up their AI deployments, Singtel has also architected the CoE to serve as a testbed that mirrors its commercial infrastructure, which Chang likened to a national power grid comprising AI datacentres acting as generators, fixed networks as transmission lines, and edge locations as substations.</p> <p>“Think about this centre of excellence for applied AI as a micro AI grid,” he said. “Not only do you experiment and solve bottlenecks, but when you go for full-scale deployment, you seamlessly flip over to the main AI grid and get the resources.”</p> <p>Marc Hamilton, senior vice-president of solutions architecture and engineering at Nvidia, described the partnership as providing the <a href="https://blogs.nvidia.com/blog/davos-wef-blackrock-ceo-larry-fink-jensen-huang/">five -layer foundation for AI deployments first mooted by Nvidia CEO Jensen Huang</a>.</p> <p>The foundation comprises physical land, power and datacentre facilities provided by <a href="https://www.computerweekly.com/news/366568455/Singtel-ropes-in-industry-partners-in-datacentre-push">Singtel’s Nxera datacentre arm</a>, followed by Nvidia’s graphics processing units (GPUs). The third layer is the broader AI infrastructure, including networking and cloud orchestration, followed by AI models, and finally, the applications.</p> <section class="section main-article-chapter" data-menu-title="Building AI applications"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Building AI applications</h2> <p>Hamilton said Nvidia plans to tap into its network of 40,000 AI startups to help Singtel’s customers build AI applications. He also stressed the importance of open models, such as <a href="https://huggingface.co/collections/nvidia/nvidia-nemotron-v3">Nvidia’s Nemotron</a>, for <a href="https://www.computerweekly.com/feature/Sovereign-cloud-and-AI-services-tipped-for-take-off-in-2026">sovereign AI</a>, ensuring that “Singapore’s data and unique competitive advantage stays in Singapore, and controlled by Singapore companies”.</p> <p>In addition, Nvidia will work with Singtel to prepare datacentres for extreme power densities without running afoul of strict new sustainability metrics.</p> <p>Singtel’s Nxera datacentres currently operate <a href="https://www.techtarget.com/searchenterpriseai/news/366574412/Nvidia-unveils-new-AI-Blackwell-chip-microservices-and-more">Nvidia GB200 Blackwell systems</a> running at roughly 200kW per rack. However, Chang noted that the CoE is already preparing for the 2027-29 deployment of Nvidia’s <a href="https://www.computerweekly.com/news/366636948/Nvidia-unveils-Vera-Rubin-architecture-to-power-AI-agents">next-generation Rubin Ultra chips</a>.</p> <p>“The next generation I’m talking about is 600kW to one megawatt per rack,” he said. “That’s insane in terms of power density – 60 to 100 times more than the average datacentre today.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI infrastructure in APAC</h3> <ul class="default-list"> <li>Nvidia has announced a slew of <a href="https://www.computerweekly.com/news/366639266/Nvidia-backs-Indias-sovereign-AI-push-with-gigawatt-scale-infrastructure">partnerships with India’s leading infrastructure and technology providers</a> to help the subcontinent build up its sovereign AI capabilities.</li> <li>A consortium led by <a href="https://www.computerweekly.com/news/366638132/South-Korea-debuts-foundation-model-in-sovereign-AI-push">SK Telecom has built a sovereign AI model</a> designed to reduce reliance on foreign tech, lower costs for local industry and propel South Korea into the top ranks of AI powers.</li> <li>Micron presses ahead with its global expansion strategy, adding <a href="https://www.computerweekly.com/news/366637651/Micron-breaks-ground-on-24bn-Singapore-wafer-fab">new NAND flash capacity in Singapore</a> to meet growing demand for memory chips fuelled by the global AI boom.</li> <li>Digital Realty <a href="https://www.computerweekly.com/news/366637290/Digital-Realty-enters-Malaysia-with-acquisition-of-Cyberjaya-datacentre">expands Southeast Asian footprint</a> beyond Singapore and Jakarta, acquiring a connectivity hub in Cyberjaya with plans for a 14MW campus to support regional AI and cloud workloads.</li> </ul> </div> </div> <p>Hamilton pointed out that this extreme density requires less datacentre real estate. Because Blackwell chips are 50 times more energy-efficient at running AI models than the previous Hopper generation, massive compute power requires vastly less physical space. “The GPU datacentre of today, versus many football fields, is much more like a basketball field,” said Hamilton.</p> <p>The efficiency of AI datacentres will be critical as the Singapore government prepares to tighten regulations on the sector. Later this year, the Singapore government is expected to table the new Digital Infrastructure Act in parliament.</p> <p>The proposed legislation seeks to establish baseline energy efficiency requirements for all datacentres, including new and existing facilities, as well as mandatory cyber security measures and incident reporting requirements for major cloud service providers and datacentres to ensure economic resilience.</p> <p>Beyond helping enterprises deploy and scale AI applications in AI datacentres, the CoE will also focus on <a href="https://www.computerweekly.com/news/366581516/Edge-AI-explained-Everything-you-need-to-know">edge AI</a> and low latency 5G networks as the industry moves beyond generative AI into <a href="https://www.techtarget.com/whatis/feature/Physical-AI-explained-Everything-you-need-to-know">embodied and physical AI</a> such as autonomous robotics, humanoids and drone swarms.</p> <p>The CoE will be located at the Punggol Digital District, which is currently being transformed into a precinct-scale testbed to help companies trial and commercialise real-world robotic applications in live operational environments.</p> <p>Chang said the CoE is expected to open its doors in about three months. In the interim, large customers in sectors such as government, healthcare, banking and transportation are already defining use cases and tapping into Singtel’s existing GPU reserves to kickstart their AI journeys.</p> </section> Singtel and Nvidia have teamed up on a multimillion-dollar facility to help organisations scale enterprise AI deployments, tackle extreme datacentre power densities, and prepare for the era of embodied AI https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/data-servers-code-datacentre-monsitj-adobe.jpg https://www.computerweekly.com/news/366639492/Singtel-Nvidia-to-help-scale-enterprise-AI-deployments Tue, 24 Feb 2026 05:00:00 GMT <p>Quantum computers are expected to become capable of breaking the cryptographic algorithms that secure the world’s digital infrastructure within the next decade. Yet, while awareness of the so-called “<a href="https://www.techtarget.com/searchdatabackup/feature/The-future-of-quantum-data-centers-Resilience-and-risk">Q-Day</a>” is high, many enterprises remain unprepared to mitigate that risk.</p> <p>According to the newly released <i>2026 Global state of post-quantum and cryptographic security trends</i> study by Entrust and the Ponemon Institute, nearly half (49%) of Singapore cyber security leaders believe a quantum computer capable of breaking <a href="https://www.techtarget.com/searchsecurity/definition/RSA">Rivest-Shamir-Adleman</a> (RSA) and <a href="https://www.techtarget.com/searchsecurity/definition/elliptical-curve-cryptography">elliptic-curve cryptography</a> (ECC) encryption will emerge within just five years.</p> <p>Despite this, just a third of organisations in Singapore are actively preparing to transition to <a href="https://www.computerweekly.com/opinion/Quantum-risk-to-quantum-readiness-A-PQC-roadmap">post-quantum cryptography</a> (PQC) – down from 36% in 2023. Globally, only 38% of organisations report actively preparing for the shift.</p> <p>“The quantum risk is no longer just a theory,” said Lawrence Tan, head of technical sales consulting for digital security for Asia-Pacific and Japan at Entrust. “These statistics are concerning because they point to the fact that awareness is high, but action is stalling. Most organisations underestimate the scale of the cryptographic transformation that’s required.”</p> <p>While mainstream commercial quantum computing remains on the horizon, the threat is already here, with state actors starting to harvest encrypted data, Tan warned. “You don’t need to wait for the quantum computer to be able to break encryption to think about migrating,” he said.</p> <p>Ssu Han Koh, solutions engineering director at CyberArk, which was <a href="https://www.computerweekly.com/news/366628313/Palo-Alto-Networks-to-acquire-CyberArk-for-25bn">acquired by Palo Alto Networks in a $25bn deal</a>, echoed this urgency, particularly regarding personally identifiable information (PII).</p> <p>“Depending on the type of data you have and how long the data is relevant, there’s a big risk because the data is still sensitive years down the road,” Koh said. “If we don’t start looking at it now, it will be a challenge to comply with <a href="https://www.techtarget.com/searchstorage/feature/Navigate-PII-data-protection-and-GDPR-to-meet-privacy-mandates">PII regulations</a>, because the data can be easily recovered or used by bad actors.”</p> <section class="section main-article-chapter" data-menu-title="Why certificate management matters"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why certificate management matters</h2> <p>For IT and security teams wondering where to begin their post-quantum journey, there’s already an immediate push to get moving: the shortening lifespan of digital certificates, which contain public keys and verify identities of websites, individuals, devices, servers, applications and services.</p> <p>Under the new CA/Browser Forum mandates rolling out from March 2026, maximum certificate validity will be progressively reduced from 398 days to just 47 days by 2029. This creates a logistical hurdle for enterprises, but it’s also the first step to achieving quantum readiness.</p> <p>But how does managing digital certificates prepare organisations for Q-Day? The link lies in the underlying cryptography. Digital certificates rely on mathematical algorithms to authenticate identities and secure communication channels. These algorithms will become obsolete when Q-Day arrives.</p> <p>And if an organisation is manually tracking its certificates, finding and replacing thousands of vulnerable keys across the entire enterprise with <a href="https://www.techtarget.com/searchcio/tip/Quantum-resistant-algorithms-Why-they-matter">quantum-resistant algorithms</a> will be an impossible task. However, by adopting certificate lifecycle management practices to handle renewal cycles, organisations can achieve what has been dubbed “crypto agility”.</p> <p>“When certificate lifespans start to get shorter, the workload is going to be many times greater,” said Koh, adding that organisations will have to embrace automation and start thinking about how they manage their certificates, including knowing what certificates they have and where they are.</p> <p>According to the Entrust study, only 43% of Singapore respondents have full visibility over their certificates, and 62% say managing cryptographic assets is extremely or very difficult.</p> <p>Tan likened the current state of certificate visibility to an unlabelled cryptography graveyard. “Without knowing when the cryptography was created, who the owners are and other attributes, you won’t know if a key was used to protect critical data or a test environment,” Tan said. “Organisations need to start labelling them to prepare for quantum migration.”</p> <p>The challenge is further compounded by the <a href="https://www.techtarget.com/searchsecurity/tip/CISOs-guide-to-nonhuman-identity-security">increasing number of non-human identities</a> and the growing use of microservices and <a href="https://www.techtarget.com/searchenterpriseai/definition/agentic-AI">agentic AI</a>, all of which require cryptographic protection.</p> <p>According to Koh, many enterprises still use manual tools like spreadsheets to manage digital certificates. While some enterprises may loosely track public certificates, they often leave private certificates, secure shell keys, and <a href="https://www.techtarget.com/whatis/definition/API-key">application programming interface (API) keys</a> entirely unmanaged.</p> <p>In particular, API keys, often generated by developers to connect to cloud services and applications, must be secured in the same way as passwords with <a href="https://www.techtarget.com/searchcloudcomputing/tutorial/How-to-perform-and-automate-key-rotation-in-Azure-Key-Vault">90-day rotation policies</a> to prevent them from becoming backdoors, Koh added.</p> <p>Indeed, achieving quantum readiness should not be decoupled from broader security practices, Tan warned. “It’s part of <a href="https://www.techtarget.com/searchsecurity/feature/How-to-implement-zero-trust-security-from-people-who-did-it">zero trust</a>, and if you don’t have post-quantum readiness, your zero-trust framework will not stand. You need to make sure your devices and your entire infrastructure are protected.”</p> <p>With the US National Institute of Standards and Technology having <a href="https://www.techtarget.com/healthtechsecurity/feature/Understanding-NISTs-post-quantum-cryptography-standards">released post-quantum cryptographic standards</a>, Koh noted that the time to build a responsive, automated cryptographic foundation is now.</p> <p>“Organisations need to be ready to be crypto agile first. That runway is shortening, and if you are not ready for it, then it can hit you quite fast,” he said.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about cyber security in APAC</h3> <ul class="default-list"> <li>Singapore mobilised over 100 cyber defenders to neutralise a sophisticated APT actor which infiltrated Singtel, StarHub, M1 and Simba networks in the country’s&nbsp;<a href="https://www.computerweekly.com/news/366638973/Singapore-mounts-largest-ever-cyber-operation-to-oust-APT-actor">largest coordinated cyber incident response to date</a>.</li> <li>Japan’s Nikkei has confirmed a major data breach that potentially&nbsp;<a href="https://www.computerweekly.com/news/366634243/Nikkei-data-breach-exposes-personal-data-of-over-17000-staff">exposed the personal information of more than 17,000 employees</a>&nbsp;and business partners after hackers infiltrated its internal Slack messaging platform.</li> <li>Australian privacy commissioner warns that the&nbsp;<a href="https://www.computerweekly.com/news/366633983/Fewer-data-breaches-in-Australia-but-human-error-now-a-bigger-threat">human factor is a growing threat</a>&nbsp;as notifications caused by staff mistakes rose significantly even as total breaches declined 10% from a record high.</li> <li>Philippine bank&nbsp;<a href="https://www.computerweekly.com/news/366633428/BDO-Unibank-taps-Zscaler-to-secure-cloud-migration">BDO is shoring up its cyber security capabilities</a>&nbsp;to protect its data and systems as it moves more services to the cloud and expands its physical presence into remote areas of the archipelago.</li> </ul> </div> </div> </section> With quantum computing threatening current encryption standards, experts call for organisations to achieve crypto agility by managing the lifecycle of certificates and cryptographic keys through automation https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/key-security-success-fotolia.jpg https://www.computerweekly.com/news/366639320/Why-crypto-agility-is-key-to-quantum-readiness Mon, 23 Feb 2026 01:06:00 GMT <p>As more consumers prepare to use personal <a href="https://www.computerweekly.com/feature/Getting-started-with-agentic-AI">artificial intelligence (AI) agents</a> to research and purchase goods on their behalf, ensuring these digital assistants don’t go rogue and embark on a shopping spree has become a security concern in the e-commerce industry.</p> <p>According to data from Akamai, traffic from AI agents and bots has nearly tripled over the past year. During a two-month period, the commerce industry saw over 25 billion AI bot requests, with China, India and Singapore accounting for most of the traffic.</p> <p>Against this backdrop, online merchants that have focused on fending off malicious bot traffic will need to rethink their security strategies, said Reuben Koh, director of security technology and strategy at Akamai. “Merchants have been so used to the old thinking that bots and automation are the enemy,” he said. “Now, they are in a split-brain scenario – they can’t block everything anymore because fewer humans are visiting their sites, but they also can’t open the floodgates.”</p> <p>To mitigate potential security risks for merchants looking to accommodate more shopping agents, and to prevent these agents from being compromised by threat actors, <a href="https://www.akamai.com/newsroom/press-release/akamai-and-visa-join-forces-to-secure-the-next-era-of-agentic-commerce">Akamai has partnered with Visa</a> to secure agentic transactions, leveraging Visa’s <a href="https://developer.visa.com/capabilities/trusted-agent-protocol/overview">trusted agent protocol</a> (TAP) alongside Akamai’s behavioural intelligence.</p> <p>“Think of Visa as the passport provider and Akamai as the immigration officer,” said Koh, adding that Visa will verify the identity of the agent and the human behind it, and cryptographically sign the intent of the transaction. Akamai then acts as border control by analysing the agent’s origin and behaviour in real time.</p> <p>“A good agent verified by Visa that comes from a bad location, such as a compromised server, will be blocked by us,” he added. “And if a verified agent enters with the stated intent to browse but suddenly begins scraping data or testing credit card numbers, Akamai’s behavioural intelligence will pick that up and stop it.”</p> <p>While Visa verifies identity and Akamai secures transport and behaviour, model creators must do their part to ensure agents do not deviate from user instructions, said Koh. “The agent’s ability to execute accurately comes from the agent provider,” he said. “Their responsibility is to <a href="https://www.techtarget.com/searchenterpriseai/tip/Why-does-AI-hallucinate-and-can-we-prevent-it">ensure the agent doesn’t hallucinate</a> – for example, if I ask it to buy a pair of sneakers, it doesn’t buy two sacks of rice because it feels I need them.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Open standards key in agentic commerce</h3> <p>Akamai’s partnership with Visa was made possible by its early adoption of the <a href="https://datatracker.ietf.org/doc/charter-ietf-webbotauth/">Web Bot Authentication (WBA) standard</a>, which, along with the <a href="https://www.techtarget.com/searchcustomerexperience/news/366638716/Vendors-hawk-everything-AI-retailers-take-measured-approach">Universal Commerce Protocol developed by Google and Shopify</a>, serves as the foundation for the emerging agentic economy.</p> <p>Because WBA is an open standard – likely to be ratified by the Internet Engineering Task Force – it prevents supplier lock-in and allows Akamai to support multiple payment ecosystems simultaneously. While proprietary layers like Visa’s TAP add specific identity features, WBA serves as the anchor standard for authenticating bots and providing additional information about their operators to websites.</p> <p>“If there’s another payment provider that comes up with their own standard that’s based on WBA, then it makes it easy and quick for us to integrate our stuff with them,” he explained.</p> <p>Koh noted that this architectural decision enables Akamai to expand its partnerships beyond its current arrangement with Visa, adding that the company is in active negotiations with other major payment providers and fintech startups.</p> </div> </div> <p>Beyond preventing errors, agent providers must also implement guardrails to prevent financial damage, such as an agent going on a spending spree or maxing out a user’s credit card. “We need to ensure the agent doesn’t go on a rampage,” said Koh. “All three of us – Visa, Akamai and the agent provider – need to work in tandem.”</p> <p>That includes securing the substrate that underpins agentic AI: the <a href="https://www.computerweekly.com/feature/Enterprise-strategies-for-API-management">application programming interfaces</a> (APIs) that agents rely on to gain access to systems and data.</p> <p>“For AI to interact with the real world – buying things, creating documents or booking airline tickets – it requires APIs,” said Koh. “But from an attacker’s perspective, I don’t need to attack the agent; I just need to target the APIs executing the instructions.”</p> <p>He warned of scenarios involving goal hijacking, where an <a href="https://www.techtarget.com/searchsecurity/tip/How-to-fix-the-top-API-vulnerabilities">attacker manipulates the API data pipeline</a> to alter an agent’s objectives – for example, changing an instruction to buy $200 worth of groceries into a transaction for a $5,000 luxury item.</p> <p>With API transactions set to balloon as automated agents begin shopping at scale, Koh believes API security will become the weakest link. “We’re already seeing API attacks on traditional applications; the growing number of agentic transactions is only going to compound the problem,” he said.</p> <p>Making matters worse is <a href="https://www.techtarget.com/searchenterpriseai/tip/Explore-mitigation-strategies-for-LLM-vulnerabilities">excessive agency</a> – one of the Open Web Application Security Project’s top 10 list of large language model security vulnerabilities – where developers, under pressure to ship products quickly, grant AI agents more permissions than necessary.</p> <p>“We are dealing with non-human identities at scale – not 20 employees, but 500,000 agents,” said Koh. “If an agent has too many permissions and suffers from hallucination or bias, the guardrails may no longer be effective.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI governance and security in APAC</h3> <ul class="default-list"> <li>Computer Weekly speaks to Keeper Security’s leadership on how identity and access management systems are becoming unified identity platforms capable of <a href="https://www.computerweekly.com/news/366638413/Interview-Why-identity-is-the-nucleus-for-cyber-security">securing both human and machine identities</a>.</li> <li>Singapore has launched a <a href="https://www.computerweekly.com/news/366637674/Singapore-debuts-worlds-first-governance-framework-for-agentic-AI">governance framework for agentic AI systems</a>, which are capable of independent reasoning and action, to address the growing security and operational risks posed by AI agents.</li> <li>Dataiku’s field chief data officer for Asia-Pacific and Japan discusses how implementing <a href="https://www.computerweekly.com/news/366637344/AI-governance-provides-guardrails-for-faster-innovation">AI governance can accelerate innovation</a> while mitigating the risks of shadow AI.</li> <li>As AI agents are given more power inside organisations, Exabeam’s chief AI officer argues they must be <a href="https://www.computerweekly.com/news/366631420/Exabeam-Treat-AI-agents-as-the-new-insider-threat">monitored for rogue behaviour</a> just like their human counterparts.</li> </ul> </div> </div> With AI agents increasingly acting as digital concierges for shoppers, verifying bot identities, securing the APIs they rely on and detecting anomalous behaviour will be key to safeguarding automated transactions, according to Akamai https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/e-commerce-security-fotolia.jpg https://www.computerweekly.com/news/366639228/What-it-takes-to-secure-agentic-commerce Fri, 20 Feb 2026 03:00:00 GMT <p>If last year proved anything in cyber security, it’s this: the battleground doesn’t just belong to those with the best tools or largest teams anymore. It belongs to those who can act faster than their adversaries.​</p> <p>2025 was a watershed year – not because attackers unveiled new zero-days or defenders fell behind – but because of one defining shift that quietly reshaped the entire landscape: speed.​</p> <p>Attackers didn’t become more intelligent. Defenders didn’t lose their skill. What changed was the tempo of operations. Intrusion, lateral movement, and exfiltration <a href="https://www.computerweekly.com/news/366627027/Digital-warfare-is-blurring-civilian-front-lines">now happen at machine speed, outpacing human escalation chains designed for a slower era</a>.​</p> <p>Organisations have spent years investing in detection fidelity and recovery strength. Now, the ones that win are those that excel in response velocity, making quality decisions with incomplete information before an attacker completes their playbook.​</p> <section class="section main-article-chapter" data-menu-title="The new attack tempo"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The new attack tempo</h2> <p>Early last year, a major Asia-Pacific logistics provider dealt with what appeared to be routine credential theft. Within an hour, the attacker had moved laterally across subsidiaries in three countries and began exfiltrating sensitive shipment data, assisted by automated tooling and reused playbooks.​</p> <p>In another case, a financial services firm in Sydney saw <a href="https://www.techtarget.com/searchsecurity/feature/Ransomware-trends-statistics-and-facts">ransomware</a> encrypt critical systems less than 90 minutes after the initial intrusion point. Their <a href="https://www.computerweekly.com/news/252468781/How-EDR-is-moving-beyond-the-endpoint">endpoint detection and response</a> (EDR) platform raised an alert within minutes, but the organisation’s escalation path required executive approval for major isolation actions. By the time that approval arrived, the attack had already propagated.​</p> <p>These examples aren’t anomalies. They’re indicators of the new norm. In 2025, reaction lag, not detection gaps, became the dominant vulnerability.​</p> </section> <section class="section main-article-chapter" data-menu-title="When certainty became a liability"> <h2 class="section-title"><i class="icon" data-icon="1"></i>When certainty became a liability</h2> <p>Cyber security principles have traditionally treated certainty as sacred: verify before isolating, confirm before containing. That approach emerged from decades of risk management discipline, but 2025 showed how dangerous it has become when timelines collapse.​</p> <p>Enterprises that insisted on confirmation before containment often discovered their confidence too late. By the time an incident was “proven,” data had already been copied, encrypted, or destroyed, and recovery options were narrower and more expensive.​</p> <p>By contrast, the enterprises that fared best weren’t reckless. They were pre‑authorised. For example, a large healthcare network in New Zealand successfully contained a stealthy persistent threat last July because of a pre‑agreed “isolate first” methodology. Their <a href="https://www.computerweekly.com/news/252478461/Inside-the-SOC-the-nerve-centre-of-security-operations">security operations centre (SOC)</a> had the authority to trigger segmented network lockdown the moment their correlation engine flagged simultaneous credential anomalies across critical systems.​</p> <p>They didn’t wait for certainty. They acted on the assumption that inaction was riskier. Later analysis showed that part of the triggering activity was benign, but leadership agreed the temporary disruption cost was trivial compared to the damage a successful compromise would have caused. The new resilience calculation: error is cheaper than hesitation.​</p> </section> <section class="section main-article-chapter" data-menu-title="The decision velocity gap"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The decision velocity gap</h2> <p>If 2024 was the year of tooling upgrades, then 2025 exposed a subtler gap: one not in technology, but decision architecture.</p> <p>Security teams today can detect faster than ever. <a href="https://www.computerweekly.com/feature/Making-sense-of-AIs-role-in-cyber-security">Machine‑learning‑based detection</a>, cross‑layer correlation, and anomaly scoring have compressed identification times to minutes. Yet, organisational latency – the delay between alert and authorisation – remains measured in hours or days.​</p> <p>The gap has become one of the most exploitable dimensions in modern defence. Attackers have no board approvals, no compliance committees, no external auditors. They can act within seconds while defenders remain constrained by governance designed for safety over speed.​</p> <p>As many CISOs across Australia and New Zealand (ANZ) are discovering, the defender’s playbook still assumes time exists. In a growing number of breaches, the adversary finishes before the defender begins.​</p> <p>Bridging the decision gap requires an enterprise to understand their minimum viable business (MVB), the smallest version of the business that can still function and serve customers when an incident compromises systems and operations.</p> <p>Rather than attempt to restore everything, everywhere, all at once, this approach focuses on essential services for revenue generation and regulatory requirements. It prioritises the minimum set of applications and data those services rely on, and the infrastructure required to run them safely, even during degraded conditions.</p> <p>In a world where attacks complete in minutes but forensics take days, understanding MVB can be the difference between business continuity and existential crisis.</p> <p>Speed changed cyber last year. In 2026, those who understand their MVB – and can restore it fast – will be the ones that stay in business long enough to tell the story.</p> <p><i>Niraj Naidu is head of engineering at Rubrik ANZ</i></p> </section> At a time when threat actors are operating at machine speed, understanding which essential systems are needed to keep the lights on can be the difference between business continuity and existential crisis https://cdn.ttgtmedia.com/visuals/German/article/business-continuity-dominoes-adobe.jpg https://www.computerweekly.com/opinion/The-year-that-speed-changed-cyber-security Thu, 19 Feb 2026 01:58:00 GMT <p>Nvidia has announced a slew of partnerships with India’s leading infrastructure and technology providers to help the subcontinent build up its <a href="https://www.computerweekly.com/feature/Sovereign-cloud-and-AI-services-tipped-for-take-off-in-2026">sovereign artificial intelligence</a> (AI) capabilities.</p> <p>During the <a href="https://www.computerweekly.com/news/366639262/India-AI-Impact-Summit-begins">India AI Impact Summit</a> in New Delhi this week, the chip giant unveiled plans to deploy tens of thousands of its latest Blackwell graphics processing units (GPUs) across the country to support the government-backed <a href="https://indiaai.gov.in/globalindiaaisummit/about-global-indiaai-summit">IndiaAI Mission that fosters AI innovation</a> by enhancing data quality, building AI talent and improving access to AI infrastructure.</p> <p>For example, Nvidia is working with engineering giant Larsen &amp; Toubro (L&amp;T) to build a gigawatt-scale network of AI datacentres. These facilities aim to keep critical data and model training within India’s borders, which is a key requirement for government and regulated industries.</p> <p>“AI is driving the largest infrastructure buildout in human history – everyone will use it, every company will be powered by it and every country will build it,” said Nvidia CEO Jensen Huang. “Together with L&amp;T – an 88-year-old engineering and nation-building leader – we are laying the foundation for world-class AI infrastructure that will power India’s growth and help realise the full vision of <a href="https://indiaai.gov.in/">IndiaAI</a>.”</p> <p>Meanwhile, cloud service provider <a href="https://yotta.com/">Yotta</a> has deployed more than 20,000 Nvidia Blackwell Ultra GPUs to power its <a href="https://shakticloud.ai/">Shakti Cloud</a>, hosted in datacentres in Navi Mumbai and Greater Noida, while E2E Networks will launch a Blackwell cluster on its AI and machine learning platform, hosted at the L&amp;T Vyoma datacentre in Chennai.</p> <p>The AI infrastructure in these datacentres will cater to the needs of model builders, startups, researchers and enterprises that are building and deploying AI applications in India, as well as supporting initiatives such as <a href="https://bharatgen.com/">BharatGen</a>, a government-backed project to create a multilingual and multimodal AI ecosystem.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> Together with L&amp;T, we are laying the foundation for world-class AI infrastructure that will power India’s growth and help realise the full vision of IndiaAI </figure> <figcaption> <strong>Jensen Huang, Nvidia</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>As part of the BharatGen project, a new 17-billion-parameter <a href="https://www.techtarget.com/searchenterpriseai/feature/Mixture-of-experts-models-explained-What-you-need-to-know">mixture-of-experts model</a> has been built from scratch using Nvidia’s NeMo framework for pre-training and the NeMo RL library for post-training. The open source model is designed to power applications across public services, agriculture, security and cultural preservation.</p> <p>In addition, Nvidia is supporting efforts by India’s $250bn IT services sector to move beyond generative AI chatbots toward agentic AI systems capable of reasoning and executing complex workflows.</p> <p>Major global systems integrators, including Infosys, TCS, Wipro and Tech Mahindra, announced they are adopting Nvidia’s AI Enterprise software to build AI agents for global clients.</p> <p>Wipro showcased a new system developed for a US health insurer that now handles 42% of inbound calls with a sub-200-millisecond latency, while Infosys has integrated Nvidia’s stack to create a 2.5-billion-parameter model for agent development, code generation, refactoring and software engineering workflows.</p> <p>“Agentic AI is reshaping India’s tech industry, delivering leaps in services worldwide,” Nvidia said in a blog post, noting that “technology leaders are tapping <a href="https://www.ciodive.com/news/nvidia-nemotron-3-power-multi-agent-systems/807952/">Nvidia Nemotron open models</a>, data and software to build and deploy <a href="https://www.techtarget.com/searchenterpriseai/tip/Agentic-AI-vs-generative-AI-Whats-the-difference">agentic and generative AI</a>, accelerating productivity and efficiency”.</p> <p>The summit also highlighted India’s growing role as a manufacturing hub, with industrial heavyweights Reliance Industries and Tata Motors adopting <a href="https://www.techtarget.com/searchcio/definition/Nvidia-Omniverse">Nvidia’s Omniverse platform</a> for building physical AI applications.</p> <p>For example, Reliance is combining Siemens’ digital twin technology with Omniverse libraries for faster, more precise simulation and plant design for its next-generation gigafactories, while Tata Motors is using <a href="https://www.computerweekly.com/feature/Urban-digital-twinsmissing-pieces-and-emerging-divides">digital twins</a> built on Omniverse libraries along with other AI capabilities to convert camera feeds into intelligent sensors for automated quality checks and real-time safety compliance.</p> <p>Nvidia said these partnerships aim to “bring physical AI into factories, warehouses and infrastructure”, allowing Indian manufacturers to reduce lead times and improve operational efficiency through high-fidelity simulations.&nbsp;</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in India</h3> <ul class="default-list"> <li>Study finds that <a href="https://www.computerweekly.com/news/366634043/Most-developers-in-Southeast-Asia-and-India-use-AI-tools">95% of developers in Southeast Asia and India use AI tools</a>, but many of their employers lack formal AI policies, leaving them to forge their own paths in upskilling and governance.</li> <li>Microsoft chairman and CEO Satya Nadella has announced a major <a href="https://www.computerweekly.com/news/366636341/Microsoft-taps-Indian-IT-giants-for-agentic-AI-rollout">expansion of the software giant’s AI footprint in India</a>, teaming up with four of the country’s largest IT services companies to deploy agentic AI capabilities across enterprises.</li> <li>Krutrim, the AI venture from Indian ride-hailing and EV giant Ola, is building a <a href="https://www.computerweekly.com/news/366629172/Olas-Krutrim-builds-AI-first-sovereign-cloud-for-India">vertically integrated sovereign cloud and AI stack for India</a> using Cloudera's data platform as a core component of its architecture.</li> <li>OpenAI is preparing a <a href="https://www.computerweekly.com/news/366630088/OpenAI-targets-India-with-datacentre-push">one-gigawatt datacentre in India</a> as part of its $500bn Stargate infrastructure push, in what would be its biggest bet yet on its second-largest user base.</li> </ul> </div> </div> Chip giant unveils compute expansion with L&T, Yotta and E2E Networks at the India AI Impact Summit, paving the way for domestic heavyweights to build AI agents and physical AI applications https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/India-trade-networking-outsourcing-adobe.jpeg https://www.computerweekly.com/news/366639266/Nvidia-backs-Indias-sovereign-AI-push-with-gigawatt-scale-infrastructure Thu, 19 Feb 2026 00:30:00 GMT 60 [email protected]