Google chief legal officer David Drummond said on Thursday that the search giant has started removing links to 250,000 web pages since a new “right to be forgotten” law was introduced in May. That’s 125,000 web pages disappearing from Google every month. About 70,000 people have requested links mentioning their names be removed, he wrote in a column for The Guardian.
The news should be shocking to anyone who cares about freedom of speech in Europe, or the freedom of anyone to do basic factual research on the internet. MarketingLand has a list of the types of stories that are disappearing from Google. Gone from the internet, bizarrely, are links to stories about Stan O’Neal of Merrill Lynch and singer Kelly Osborne.
Google opposed the new law mandating that links to stories be removed even if they are true, but now says it must abide by the rule. As Google’s online form for link removal requests shows, the standard for deleting links is not whether the web pages are accurate or true. It’s much more vague and subjective that that. The standard is: “inadequate, irrelevant or no longer relevant, or excessive in relation to the purposes for which they were processed.” The law doesn’t apply to the web pages themselves, only search engine links to them.
While the links may be upsetting for some of the people mentioned in those stories, hiding researchers’ ability to locate the stories through Google is akin to removing the card catalogue system from a library while leaving the books in place. The removal process is haphazard and imperfect, Drummond says:
So we now have a team of people reviewing each application individually, in most cases with limited information and almost no context.
Danny Sullivan of Search Engine Land has a great dissection of just how haphazard this process can be — and how it can make legitimate information that no one is complaining about suddenly disappear from Google. If two people have the same name, then links to both parties can disappear. And it’s not just name searches, it’s links to other words associated with those names that are vanishing too, Sullivan says:
To understand, let’s assume there’s someone named “Emily White” who doesn’t like that a search for her name on Google brings up information about her having gone bankrupt.
White, who let’s say lives in London, makes a request for Google to drop the links under the EU’s Right To Be Forgotten mandate. Google grants this. As a result, the links would no longer show up for these situations:
1) Searches for just “emily white”
2) Searches for her name plus other words, such as “emily white bankrupt” or “emily white london”
Any search involving her name along with other words would have the forgotten links filtered out. White wouldn’t have to provide a list of descriptive terms to go along with her name. Just the presence of her name in a search would be enough to trigger the filtering.
Already, Google has admitted it has made some mistakes. Six links to articles in The Guardian were removed and four were later restored.
Sullivan has deduced that this list of factors is being used by Google to decide which links get thrown into the memory hole:
- Does information involve public figures?
- Does information come from a reputable news source?
- Is the information recent?
- Does the information involve political speech?
- Does the information involve professional conduct?
- Does the information involve criminal convictions where time is still being served?
- Is the information published by a government?
On its own, the third criterion should be chilling: “Is the information recent?” It suggests that vast chunks of the past, no matter how crucial or interesting, will suddenly become much more difficult to locate.
Who benefits? Here’s one example of the type of person who might be quite happy to see links to the past erased: There are 20 senior politicians and government officials in Britain who are current under investigation for their alleged links to a ring of child abusers in the 1980s and 1990s.