Don’t Force Google to ‘Forget’

CAMBRIDGE, Mass. — THE European Court of Justice ruled on Tuesday that Europeans have a limited “right to be forgotten” by search engines like Google. According to the ruling, an individual can compel Google to remove certain reputation-harming search results that are generated by Googling the individual’s name. The court is trying to address an important problem — namely, the Internet’s ability to preserve indefinitely all its information about you, no matter how unfortunate or misleading — but it has devised a poor solution.

The court’s decision is both too broad and curiously narrow. It is too broad in that it allows individuals to impede access to facts about themselves found in public documents. This is a form of censorship, one that would most likely be unconstitutional if attempted in the United States. Moreover, the test for removal that search engines are expected to use is so vague — search results are to be excluded if they are “inadequate, irrelevant or no longer relevant” — that search engines are likely to err on the safe side and accede to most requests.

But the decision is oddly narrow in that it doesn’t require that unwanted information be removed from the web. The court doesn’t have a problem with web pages that mention the name of the plaintiff in this case (Mario Costeja González) and the thing he regrets (a property foreclosure); it has a problem only with search engines that list those pages — including this article and possibly the court’s own ruling — as results to a query on the basis of Mr. González’s name. So nothing is being “forgotten,” despite the court’s stated attempt to protect such a right.

How an individual’s reputation is protected online is too important and subtle a policy matter to be legislated by a high court, which is institutionally mismatched to the evolving intricacies of the online world.

Progress has been limited perhaps by a shortfall of imagination by Google, Microsoft’s Bing and the handful of other powerful intermediaries who stand between what we ask and what we’re told is relevant. Search engines generally treat personal names as search terms like any others: Data is data. Google and company have not internalized just how significant that first page of search results has become to someone whose name has been queried. What they place on that page may do more than anything else in the world to define a stranger in others’ estimations.

What if search engine companies were to think more creatively about how such searches might work? In 2007, Google admirably experimented in this area, introducing a feature to its Google News aggregator that allowed people quoted or mentioned in a news article indexed by Google News to add a comment next to that article in the search results. Such participants could offer readers of Google News an explanation, an apology, or a reason to discount whatever it was they were about to read. (Academics were among the first users of the feature, often adding a comment to contextualize something a newspaper reporter had quoted them as saying.) But Google ultimately abandoned the feature.