Right to be forgotten: having search results about your name removed

The right to be forgotten is the right to have personal data erased, laid down in Article 17 of the GDPR. Applied to search engines, it means this: Google removes the link to a page from the results that appear when someone searches for your name.

Two things are often confused. The article itself stays online; only its findability through your name disappears. And the removal applies to the European versions of the search engine, not worldwide.

The right emerged in 2014, when the Court of Justice of the European Union ruled in the Costeja (C-131/12) case that a search engine is a controller of the personal data in its results, and can therefore be held to account. It has since been developed further in cases including GC and Others (C-136/17) on sensitive data and criminal information, and TU and RE v Google (C-460/20) on inaccurate information.

The right to be forgotten is a personal right: it applies to natural persons. A company cannot invoke it, not even for a trade name. A company can still take action against harmful search results, but through other routes: copyright, the policies of the platform where the publication appears, and suppressing results with its own publications. Those routes are set out on the page remove negative search results.

For public figures the bar is higher, because the public has a greater interest in being able to find information about them. Higher is not insurmountable: what counts is the role you held, how long ago that was, and whether the result still says anything about it.

For each link, Google weighs your privacy against the public's interest in finding the information. In practice, these points carry the most weight:

• Is it accurate? Demonstrably inaccurate information weighs heavily. Since the TU and RE judgment of 8 December 2022, however, you have to show the inaccuracy yourself, with evidence that you submit along with the request.
• How old is it? A report from ten years ago about a matter that has been settled weighs differently from last month's news.
• What happened afterwards? A decision not to prosecute, an acquittal, a withdrawn investigation or a correction changes the balance, even if the original report was factually correct when it appeared.
• What is your role? A director of a listed company is in a different position from a private individual who was in the news once.
• What kind of data is it? Criminal data and special categories of personal data are subject to a stricter regime, and that works in your favour.
• Does the information remain available elsewhere? Removal from the search results for your name leaves the publication itself untouched; that makes the intervention smaller and the chances greater.

Success rates quoted without reference to the category say little. Our own reading of Google's Transparency Report (11 September 2026) shows how strongly the outcome varies by type of content. For the European Union, over the second quarter of 2026:

That last row matters: content you once published yourself is the largest category, at 1.65 million URLs, and Google removes none of it. There the route is a different one: approaching the source.

For the Netherlands, the total came to 100,069 requests covering 378,750 URLs, of which 53.5% were removed. Well over nine in ten requests came from private individuals.

Anyone can submit a request: Google provides a web form for it. What makes the difference is not the form but what you put in it.

For each URL you substantiate four things: which personal data appear, why they may not (or may no longer) be processed, what is factually inaccurate or outdated about them, and what the harm is. Since the TU and RE judgment, evidence is no longer an attachment but the core of the request: include the decision not to prosecute, the judgment, the report or the correspondence that shows what you claim.

Where it goes wrong in practice: one request for ten URLs at once with a single general justification, an appeal to emotion without a legal basis, or a request that draws attention to a matter that was fading away on its own. Sometimes, therefore, our advice is not to submit a request at all, or not yet.

A rejection is not the end of the road, and in our practice it is usually the beginning.

Mediation by the Dutch Data Protection Authority (Autoriteit Persoonsgegevens). The Dutch supervisory authority can mediate if a request has been wrongly rejected. We then submit the case again, with the same substantiation plus a specific argument as to why the rejection does not hold. The Dutch DPA assesses the merits and asks Google to respond. This route is underused in the Netherlands and, in our work, one of the most effective.

Mind the six-week deadline. Both for involving the Dutch DPA and for filing an application with the court, a deadline of six weeks applies, counted from the response to your request. A rejection you leave lying for six months costs you those routes. If you are too late, a new, better-substantiated request to Google remains an option, but that is a weaker starting position.

Going to court. If mediation does not help, an application procedure remains: against Google, against the publisher of the publication, or against the person who posted it.

Our founder has litigated against Google over search results several times; two examples are the decision of the District Court of Amsterdam of 19 July 2018 and the decision of the Amsterdam Court of Appeal of 7 April 2020.

A request to Google usually produces a first decision within 4 to 8 weeks. Mediation by the Dutch DPA adds 2 to 6 months to that. Legal proceedings take 6 to 18 months.

A right-to-be-forgotten request with us costs from €895 excl. VAT, in practice €895 to €1,495 depending on how complex the case is. Follow-up requests, for example to another search engine or for a new URL, cost less because the substantiation is already in place. Mediation and suppression are charged by the hour, at our standard hourly rate of €230 excl. VAT. Legal proceedings are in the order of €15,000 excl. VAT.

It does not take the publication down. If you want the article itself to disappear or your name to be taken out of it, the publisher is the party to talk to, not the search engine. Anonymisation is often achievable where removal is not.

It does not work outside Europe, it does not automatically cover images, autocomplete suggestions and "People also ask" boxes (separate procedures exist for those, see the main page), and it does not apply to inaccurate statements about you by AI services. For those, the route runs through the provider of the model, which is a practice of its own: see AI reputation management.

And it does not work on what you once published yourself. That is not a legal technicality but the largest block in the figures above.