The future isn't just AI — it's trillions of AI agents collaborating across the open web securely.
​​​Project NANDA (Networked AI Agents in Decentralized Architecture) which originated at MIT aims to dismantle the four critical choke points within this ecosystem's infrastructure by focusing on consensus driven innovation: 'DNS', 'CA', 'Orchestration', and 'Attestation'.
​​​Project NANDA is pioneering the foundational infrastructure for the Internet of AI Agents — enabling autonomous systems to interact, transact, and collaborate seamlessly.
​​​​​​Project NANDA employs a three-pronged approach:
(i) Technology: Building technology, standards, and reference implementations through a collective of top minds.
(ii) Social Mission: Working to maintain an open Agentic Web for innovators while ensuring safety for vulnerable populations.
(iii) Venture Ecosystem: Fostering a wide coalition of founders and investors.
Project ​​​​NANDA's development draws on proven MIT research in AutoML, split learning, privacy-preserving architectures, and decentralized coordination. Early partners in the coalition gain preferential access to emerging standards, reference implementations, and the academic-industry consortium that will influence the emerging frameworks globally.
- 6000+ Community Members
- 61 Collaborations
- 15+ Research Papers
- 100+ Industry Speakers
- 75+ Active Projects
- 18 Global Locations
Our research outlines the architectural and strategic foundations of the Agentic Web.
| Paper Title | Description | Links |
|---|---|---|
| Beyond DNS: Unlocking the Internet of AI Agents | The foundational paper. Details the design of the NANDA Index, AgentFacts schema, and Verified Agent Discovery. |  |
| Upgrade or Switch: The Need for New Registry | Analyzes why DNS fails for AI agents and weighs incremental upgrades against purpose-built registries. |  |
| NANDA Adaptive Resolver | A dynamic microservice architecture for agent name resolution in distributed environments. |  |
| NANDA in Practice: Enterprise Perspective | Explores Zero Trust Agentic Access (ZTAA) and governance for enterprise agent ecosystems. |  |
| Survey of AI Agent Registry Solutions | Compares NANDA, MCP, A2A, and MS Entra across security, scalability, and maintainability. |  |
Join our Writing Group to contribute to upcoming research.
Focus: Onboarding, Indexing, & Discovery
- Index Infrastructure: NANDA Index for agent discovery.
- Cross-Platform Bridge: Protocol bridges between A2A, MCP, and HTTPS.
- Agent Onboarding: SDKs for easy agent deployment.
Focus: Knowledge Pricing & Edge AI
- Knowledge Pricing: Mechanisms for agents to value/exchange info.
- Edge AI: Distributed intelligence at network edges.
- Economic Protocols: Incentive systems for agent services.
Focus: Large Population Models & Co-learning
- LPMs: Collective intelligence from agent populations.
- Collaborative Learning: Privacy-preserving co-learning.
- Cross-Silo Coordination: Agents working across data boundaries.
- Forbes: Line Up and Identify Yourselves — AI Agents Get Organized with NANDA
- Forbes: Make a Decentralized Internet with AI — NANDA Is Coming
- LinkedIn: NANDA: The Internet of AI Agents (Ramesh Raskar)
| Intro to NANDA | NANDA Demo |
|---|---|
 |
 |
We believe the web must evolve from static content to autonomous actors. Join us in building the next layer of the web.
- Discord: For technical discussions and community chat.
- Events (Lu.ma): Join our weekly webinars and summits.
- GitHub: Explore our code and repositories.
- FAN Fellowship: Applications are open.
- Tresata Fellowship: Apply for paid roles.
- Radius Fellowship: Apply for paid roles.
Made with ♥ by the NANDA Community
Building the Open Agentic Web