The right to be forgotten lets people in Europe ask a search engine to stop showing specific results for searches on their name. It rests on Article 17 of the GDPR, applies to searches made from Europe, and leaves the page itself published. As of November 2026, the Google, Bing and Brave indexes also feed major AI assistants, so delisting now changes what those assistants can retrieve about a person.
Last updated: November 2, 2026
What are the key takeaways on the right to be forgotten?
The right to be forgotten has three points that matter most for anyone in Europe whose name is tied to a damaging search result.
- Under Article 17 of the GDPR, a person can ask a search engine to stop showing specific results for searches on their name.
- Google, Bing and Brave are no longer just search engines: as of November 2026, their indexes also feed the major AI assistants, directly or through services built on them.
- After a refusal, the data protection regulator sided with the applicant in more than 80 percent of the complaints MediaMaze brought, according to its founder.
Why is the European route to reputation repair so often overlooked?
The European route is overlooked because most reputation advice follows American logic: if something damaging ranks for your name, you push it down by publishing better content, building profiles, building links and earning coverage. The content route has its place, and the Everything-PR guide to AI reputation management covers it.
In Europe there is a second route. You can ask the search engine to stop showing specific results for searches on your name. That route is Article 17 of the General Data Protection Regulation (GDPR), the right to erasure, which in search is usually called delisting, de-indexing or the right to be forgotten (in Dutch: recht op vergetelheid).
At MediaMaze, the Dutch reputation firm I founded in 2012, we have handled more than 600 of these requests for executives, entrepreneurs and private individuals. I have also acted as authorised representative for clients in a series of court proceedings on this subject, at the District Court of Amsterdam and the Amsterdam Court of Appeal. In this piece I explain how delisting works, why it is no longer only about Google, and why it has become relevant for what AI assistants say about people.
What is delisting under the right to be forgotten?
Delisting under the right to be forgotten means that a search engine stops showing a specific URL for searches on a person's name. Since the ruling of the Court of Justice of the European Union in Google Spain (2014), a search engine is a controller of personal data in its own right: the party responsible for what it does with that data.
A name search compiles a profile of that person from pages across the web, and the person concerned can object to the contents of that profile. The page itself continues to be published, and the publisher is rarely notified of any action taken.
What changes is the link between the name and the page. When a prospective client, employer or investor types in your name, they no longer see the story from eleven years ago that portrays you in such a wrong fashion. This delisting applies to searches on your name from Europe, not worldwide.
Which search results qualify for delisting?
A search result qualifies for delisting when the search engine finds that the privacy interest of the person outweighs the interest of the public in finding the information through a search on that name. For each request, the search engine weighs four main factors.
| Factor | How the search engine weighs it |
|---|
| Accuracy | Manifestly inaccurate information weighs heavily in favour of delisting. Since the TU and RE v Google ruling (2022), relevant and sufficient evidence is enough, and no court ruling is needed. |
| Time | A bankruptcy, a dispute or a conviction from long ago can lose its news value, while the damage of seeing it next to your name continues. |
| Role in public life | A politician or a well-known executive has to accept more than a private person. |
| Nature of the information | Data about health, sexual life or religion falls under a stricter regime. So does data about criminal convictions and offences under Article 10 GDPR, as the Court confirmed in GC and Others (2019). |
By contrast, news about professional misdeeds weighs more heavily toward the interest of the public than purely private information. A journalistic source does not exclude delisting, because the publication and the listing by the search engine are assessed separately. Journalistic sources are, however, assumed to have some public relevance.
What makes a right-to-be-forgotten request succeed?
A right-to-be-forgotten request succeeds when it is argued per URL instead of filed as a form. The most common mistake is treating the request as a form when it is an argument.
For each URL, the search engine needs to understand what the page says, why it harms you now, and why the public interest in finding it through your name is weak today. Whether a page counts as "outdated" or "irrelevant" in Google's eyes is not something you can read in the law. An expert knows it from hundreds of earlier decisions, including which URLs are better left out, because weak URLs do not make a request stronger.
What can you do when Google refuses a delisting request?
When Google refuses a delisting request, you can ask the data protection authority in your country to look at the case. In the Netherlands that is the Dutch Data Protection Authority (Autoriteit Persoonsgegevens). The regulator does not decide in the search engine's place, but it can assess whether the refusal holds up and ask the search engine to reconsider.
In my experience, this is one of the most effective routes there is, and it is underused. When a well-argued request, drafted by a qualified lawyer or legal professional, is refused and then put before the regulator, the regulator sides with the applicant in more than 80 percent of the cases we bring and asks Google to reconsider its earlier decision.
In a large share of those cases Google then reverses its refusal and delists the result after all. That is a big part of why our success rate is around 90 percent when requests, mediation by the regulator and proceedings are combined.
Should you write your request with AI?
You should not leave a right-to-be-forgotten request to AI alone, because search engines and regulators increasingly receive requests that were obviously written by AI. Those requests are long and generic, full of legal phrases that do not quite fit the case, and sometimes contain references that are simply wrong. The people who read them notice immediately, and with that notice the little time and attention that went into the request. It is hard to expect someone to go the extra mile for you when you have visibly not tried very hard yourself.
So make sure your requests, to Google as much as to the regulator, are well argued, no longer than necessary, well written and free of nonsense. AI can help enormously with drafting and checking, but in the end you are writing for a person. When you have handled many of these requests, you get to know the people who read them, sometimes personally. You take into account who they are and what they need to read in order to decide, and you structure your request accordingly.
AI cannot do this for you for a second reason: the requests and decisions are confidential. Apart from a handful of anonymised examples in Google's transparency report, they are not published, so they were never in any training data. An AI assistant knows what Article 17 says, not how Google, Bing or a regulator decide, nor what requests they receive.
Is a request to Google enough?
A request to Google alone is not enough, because Bing and Brave Search each run their own request process. Bing has its own request form and its own reviewers, and Brave Search runs an independent index and handles requests itself. A decision by Google does not bind Bing or Brave: each search engine has to be asked, and convinced, separately.
The other search engines used to be an afterthought when it came to the right to be forgotten. Because of AI, they no longer are.
Why do search engines matter for AI assistants?
Search engines matter for AI assistants because many assistants answer questions about a person by searching the web, not from memory alone. As of November 2026, the connections between the three indexes and the major assistants look like this.
| Search index | Assistants that build on it, as of November 2026 | Request process |
|---|
| Google Search | Gemini, Google's AI Overviews and AI Mode | Google's delisting form |
| Bing | Microsoft Copilot, and ChatGPT's search, which has drawn on Bing among other sources | Bing's own request form and reviewers |
| Brave Search | Claude's web search, which has been reported to use Brave | Handled by Brave itself |
So the three indexes behind the right to be forgotten also sit, directly or through services built on them, behind a large part of what AI assistants retrieve. When a page is delisted for a name at Google, Bing and Brave, it stops coming back when an assistant searches on that name from Europe. That prevents the page from being read, summarised or cited in that answer. That is why every request we handle now goes to all three.
There are limits to this effect. A model also draws on its training data and on its provider's own crawlers. For those layers the AI providers themselves are the right address, under a separate right that I cover in a follow-up piece. For how retrieval choices shape which brands and people appear in answers, see what generative engine optimization (GEO) covers, and browse Everything-PR research on how AI assistants cite sources.
What should you do if a search result harms you?
A person in Europe whose name is tied to a harmful search result should first answer three questions about that result.
- Is the information inaccurate?
- Is it outdated?
- Is it about your private life rather than a public role?
The difficulty in this assessment is that the honest and accurate answer is often not the one you would give yourself. Someone who has seen hundreds of decisions knows whether your case meets that bar, and how to argue it at Google, Bing, Brave and, if needed, the regulator or the courts. Executives who want to decide which results to address first can start with the executive reputation management guide, and communications teams can use the corporate reputation management guide.
