DUBLIN -- Much of the coverage of the European Union's so-called right to be forgotten ruling focuses on news stories de-indexed from Google. But the biggest source of material that has been subject to requests isn't news at all. It's public Facebook posts indexed in search.
"We're getting a huge number of requests from Facebook," said Peter Barron, Google's head of communications for Europe, the Middle East and Africa, as part of a panel sponsored by the Guardian Newspaper at the Web Summit in Dublin, Ireland on Wednesday. Most of these requests could be handled if the user simply understood their Facebook privacy settings.
So far, Google has received 160,000 requests spanning 500,000 separate URLs, and the search engine is reviewing each by hand, a labor-intensive process involving "scores" of staff attorneys as well as contractors hired for that purpose. "We've taken the view right from the beginning that every case requires human judgment and intervention," Barron said. "Our view is we need human oversight. Some require a huge amount of thought based on not very much information."
The vast majority of cases are relatively straightforward. As it stands, Google is complying with 58 percent of requests, while 42 percent are denied.
But the panel painted a nuanced picture of the ruling, which has received heavy criticism from Silicon Valley tech companies as well as media organizations in the U.S. Before the ruling, anyone in Europe who wanted something de-indexed had to take it up with the publisher of the content. But the ruling reclassified search engines as data controllers, which could apply to any search engine -- it's just that Google happens to have 90 percent of the search market.
"We felt individuals had the right to approach a search engine in these cases," said Steve Wood, head of policy delivery in the Office of Information Commissioner. "It could apply to anyone indexing the Internet, not just Google."
But having complied with the ruling since May, both Google and personnel at the Guardian Newspaper have been struck by how few requests are from subjects of news stories; rather, they tend to be individuals only obliquely connected to an issue. But the mere presence of their name makes them collateral damage in search results.
"What's interesting is how few of these are actually about press articles, and if they are, they are about incidental mentions," said Guardian journalist Julia Powles. The Guardian has been notified of 29 separate removals connected to 24 different people. Of those, four have been reinstated.
Because Google doesn't disclose who asked for a link to be removed, it is often difficult to tell who made the request and why. For example, Google removed an unflattering story written by BBC economics editor Robert Peston about Stan O'Neil not because the former Merrill Lynch boss requested it but because a commenter thought better of something he'd written.
In another case, a story was removed about an 11-year-old girl killed by a pedophile, at the request of someone who was quoted in the story and was disturbed by having that link surface on the front page in searches conducted on their name.
Powles brought up the case of a 16-year-old bankruptcy notice that for some reason remained very prominent in searches for an individual's name many years after it was resolved.
Most of the approvals are for individuals with relatively small digital footprints, for whom a single link could have a big impact. "When you have a prominent person and there are a lot of articles on them, there is forgetting," Barron said. "It is much tougher if there is someone one person linked to them in their lifetimes."