One aspect of the Google Spain judgment I’ve not seen discussed is the incentives it creates for search engines. The European Court of Justice found that under some circumstances Data Protection law entitles an individual to demand that out of date and inaccurate results be removed from the results of a search for their name (the unusual circumstances of the case meant it wasn’t possible to remove the out of date information from the original website). If a search engine fails to remove the link then an appeal can be made to the appropriate national data protection regulator or court. If the search engine does remove the link, even if the relevant circumstances do not apply, then as far as I can see there’s no risk of any legal case being brought. Under UK law, at least, there’s no right to be linked nor does anyone have a contract with a search engine that might be breached by the removal of links. So when a search engine receives a request to de-link, it has the choice between an action that might result in a further legal case and one that definitely won’t. Which would you choose?
As the Court’s own adviser noted (at para 133), we’ve been here before. The eCommerce Directive’s rules on intermediary liability create the same one-way bet. When a hosting service receives an allegation that third party material is defamatory or infringes copyright or other law, the host risks liability if it continues to serve the material but none if it removes it. For hosting services there may be a contract with the provider of the information but most of those contain an exclusion from liability when a complaint is made. A research project by Bits of Freedom indicated that most hosts did indeed remove material without full investigation of copyright complaints; the Electronic Frontier Foundation suggest that both copyright investigators and hosting companies have automated processes that follow the incentives the law (in that case US law) has created.
I had expected that Google, having lost their ECJ case, would do the same and simply remove any link for which a request was received. The business case for doing so looks clear. That would be particularly damaging for anyone whose business depends on them being found by search – suppose a competitor made an unsubstantiated request to de-link? Such claims have been made in the past against other intermediaries under copyright and defamation law. Not only do search engines have no way to verify that the person making the request has the authority of the subject of the link, the incentives the law has created give them no reason to even try.
In fact, Google don’t seem to have taken that easy way out, but have been publicly raising the issues around the judgment. No doubt they have commercial reasons for doing so: perhaps they consider that deleting links without question would damage users’ confidence in search results. But I’d much rather have public discussion of what society actually needs of Internet intermediaries (for example in articles by Julia Powles and Luciano Floridi on the balance of rights and jurisdiction issues) than have intermediaries silently decide that themselves.
UPDATE: Andres Guadamuz (technollama) has pointed out that examples in Google’s transparency report suggest that they are indeed taking more care over requests than the legal incentives would indicate. And a Dutch court hasn’t blindly followed the ECJ’s simplistic suggestion that the right merely depends on whether or not the person asking to be de-linked is “famous”. Both good news, and I hope examples that others will follow.
