Right to be forgotten, one year on: Google processes nearly one million links

Google has received around a quarter of a million removal requests by Europeans since a ruling from last May forced the search giant to accept them.
Written by Liam Tung, Contributing Writer

Europe's right to be forgotten ruling is nearly one year old. Since the ruling was handed down last May, EU residents have asked Google to remove just under one million links from its results.

The European Court of Justice's landmark privacy ruling in 2014, known as the 'right to be forgotten', made search engines responsible for the data they process. It stipulated that individuals could ask for search engines to remove certain links from the results of searches for their names, if the links lead to material that is out of date, excessive, or irrelevant.

Google said on Tuesday it has received just over 250,000 requests from Europeans asking the company to review 920,258 links from search results since the ruling came into effect.

Google's dominance of the search market in Europe has left it shouldering the bulk of requests. It must decide whether the requests are valid or not according to whether it views the content linked to as "inadequate, irrelevant or no longer relevant, or excessive in relation to the purposes for which they were processed and in the light of the time that has elapsed".

If people disagree with Google's decision, they can refer their request to a national data protection authority, which can then take the matter up with Google to come to a final decision.

According to Google's transparency report, it has now removed 321,543 URLs from search results for people's names, representing 41.3 percent of all requests evaluated. However, its report doesn't include the number of requests still under review. Figures quoted by the Wall Street Journal today indicate 15 percent of the links are still under review.

At a conference in Berlin this week, Google's global privacy counsel Peter Fleischer gave some indication of how the company makes decisions, claiming it is doing its best to play a role "we never asked to play - and don't want to play", the WSJ said.

Requests, whether submitted through Google's online form or by mail, first go to a large team of lawyers, paralegals, and engineers, according to Fleischer. That amounts to "dozens" of people, mostly working in Google's Dublin office.

All cases go to a vote by the Google employees, and while easy cases are dealt with swiftly, harder ones are referred up to a senior Google panel, sometimes with the help of external lawyers specialising in the field.

According to Fleischer, Google often reaches consensus on decisions but on some cases "we are violently split".

Forget.me, a website that submits requests to all search engines on behalf of users, says the number of URLs sent to Google for review has tapered out in recent months.

In the first three months, it was submitting an average of 1,500 requests per day, which has now declined to 500 per day. Since June 24, it's sent 61,753 URLs to Google.

Google's own figures also indicate a slowdown in the number of requests it receives. In the first six months of taking requests, it received 144,907, compared with 108,710 in the most recent six months. It assessed nearly 500,000 URLs in the first six months, and a further 420,000 URLs in the second six months.

Google has become faster at processing requests, according to Forget.me's figures: in June 2014 it took 56 days to process a requestion, but by March this year that figure fell to 16 days.

Google is also saying 'no' more often. In June last year it declined 43 percent of requests, but since October it refused around 70 percent of requests.

The vast majority of requests (58.7 percent) concern "invasion of privacy", according to Forget.me, while "damage to reputation" accounts for 11.2 percent of requests.

The most commonly refused requests are those labelled as "concerns your professional activity", which make up 26 percent of the total.

Read more on the right to be forgotten

Editorial standards