Remove an article from Google
By Yair Cohen, Solicitor specialising in the right to be forgotten
Google can be required to stop showing an article about you when someone searches your name. Lawyers call it delisting, and it is a legal right rather than a favour. Article 17 of the UK GDPR gives you the grounds, Google decides in the first instance, and if Google turns you down the decision does not stand unchallenged. Our solicitors have been making delisting applications for over two decades, covering news reports, court reporting, blog posts, forum threads, directory entries and old professional profiles.
Delisting does not delete anything. The article stays where the publisher put it, and what changes is that your name stops leading to it. For most people that is where nearly all of the damage sits, because almost nobody finds these stories any other way. If you want the article taken down at source as well, that is a different application to a different target, and how to remove news articles from the internet explains it.
This article is about the Google side: what Google is actually deciding, how to read your own chances before you apply, what the application needs to contain, and what to do with a refusal.
The decision Google is actually making
Google is not deciding whether the article was fair, whether the journalist behaved well, or whether the story ruined your life. It is deciding one narrow question: does this link still earn its place in the results for your name.
That framing surprises people, and getting it wrong is the most common reason a self-made application fails. Applications that describe how much distress the article caused tend to come back refused. Applications that show the story no longer tells anyone anything useful about who you are today tend to succeed.
The reason Google has to decide at all goes back to Google Spain SL v AEPD and Mario Costeja González (C-131/12), which established that a search engine indexing pages about you is processing your personal data in its own right. Google is a data controller, not a neutral pipe, and it carries a controller's duties towards you.
There is a further point that people miss, and it does more work than anything else on this article. In NT1 and NT2 v Google LLC [2018] EWHC 799 (QB) the High Court held that Google could not rely on the journalism exemption for its own processing. The exemption protects the newspaper's decision to publish. It does not protect Google's decision to keep pointing at it. That is why a delisting application can succeed even where the newspaper is perfectly entitled to keep the story on its website, and it is why the two routes have different answers as well as different targets.
Google is not doing you a favour when it delists an article. It is complying with a duty it has owed you since 2014.
The four things that decide your case
You can read your own chances reasonably well before anyone drafts anything. Four factors carry nearly all the weight, and NT1 and NT2 is the clearest illustration of them, because the same judge granted one claim and refused the other on exactly these grounds.
How long ago it was
Time is the strongest single factor. A story from last year is close to unarguable. A story from a decade ago, about something that has not repeated, is close to unanswerable the other way. Everything in between turns on the other three.
What you do now, and what you did then
Google reads public role broadly, and that catches people out. It is not limited to politicians and celebrities. Company directors, regulated professionals, anyone who trades on their own name, and anyone whose current work connects to the subject of the story will find Google weighing the public's interest more heavily. Where you have moved away from the world the story belonged to, say so and evidence it.
What the article is about
Some subject matter carries extra protection rather than less. In GC and Others v CNIL (C-136/17) the Court of Justice held that where an article reveals special category data, which includes health, sex life, religious or political beliefs and criminal offence data, delisting should ordinarily follow unless a strict public interest reason keeps it. A large share of the applications we make sit squarely in that category, and most people making their own applications do not realise it.
Whether it is accurate
Inaccuracy is an extra ground rather than a required one, and it comes with a burden. In TU and RE v Google LLC (C-460/20) the Court of Justice held that a person alleging the indexed content is inaccurate should produce the evidence reasonably available to them, and that Google is not expected to investigate the truth of a story itself. That judgment postdates Brexit, so it does not bind the English courts, although they may have regard to it. Either way the practical lesson holds: an inaccuracy point without documents attached rarely moves anyone.
Why your case may be stronger than you think
People arrive convinced they have no case, usually for one of three reasons, and all three are misconceptions.
The first is thinking the article has to be untrue. It does not. The question is whether continued listing of accurate, lawfully published reporting is still justified years later, and Google Spain settled that it may not be.
The second is thinking a conviction closes the matter. The Rehabilitation of Offenders Act 1974 does the opposite once the conviction is spent, because the law then treats you as rehabilitated, and criminal offence data is also special category data under the GC and Others reasoning. Those two points stack, and spent-conviction applications are among the strongest we make.
The third is thinking that being charged, or arrested, or named in proceedings that went nowhere, counts against you. It is the reverse. An acquittal, a discontinuance or dropped charges leave the story reporting something that never became a fact about you, and that is exactly the position Article 17 exists for. Article 21 of the UK GDPR adds a right to object to the processing outright, which we usually plead alongside.
What a delisting application needs to contain
A delisting application is a case file rather than a form, and the difference shows in the outcome.
A complete schedule of addresses. Google decides address by address, so a story that appears on five websites needs five separate decisions, and Google can agree to some and refuse others. The schedule covers the original article, the syndicated copies, the aggregators, the scraper sites, the image results and any page quoting the story. Leaving one out leaves your name attached to it.
The grounds, taken URL by URL. Article 17 and Article 21, with the authorities, applied to each address rather than asserted in general. Two URLs carrying the same story can attract different answers depending on how each one is framed.
Evidence of what has changed. This is the part people leave out. The date the conviction became spent, the acquittal, the discontinuance letter, the change of career, the years without repetition. Assertion does not carry an application; documents do.
A position you can stand behind. An application commits you to a factual account, and a refusal is harder to reopen than a first application is to get right. That is the reason a specialist should draft it, and the reason we assess the case before anyone writes anything. The assessment costs you nothing.
Once it goes in, Article 12(3) of the UK GDPR requires Google to tell you what it has done without undue delay and within one month, which it may extend by a further two months where the application is complex. In our experience most decisions arrive inside that window; applications touching court reporting take the longest and sometimes attract questions first.
On fees, the variables are the number of publications involved and whether the matter resolves at the application stage. In most cases we act on a fixed fee, agreed with you in advance, which depends on the complexity of the case and the number of publications involved. The initial expert assessment of your case is free, and you will know where you stand before committing to anything.
Where delisting stops
Delisting is powerful and it is not total, and knowing the edges before you start avoids disappointment later.
It is tied to your name. Delisting takes the article out of the results for searches of your name, and someone searching the subject matter can still reach it. That covers the way nearly everyone actually looks you up, which is why the remedy works, but it is not invisibility.
It is tied to Google. Bing keeps its own index, takes its own applications and feeds other services, so we apply in parallel rather than afterwards.
It is tied to Europe. In Google LLC v CNIL (C-507/17) the Court of Justice held that a search engine is not required to delist across every version of its service worldwide. Geographic filtering covers most real-world searches about you, and a determined searcher abroad may still find the page.
It leaves the article live. The publisher's page stays up, and anyone with the address can read it. Where the story matters enough to pursue at source as well, how to remove news articles from the internet sets out that route, and the two work best planned together.
Two loose ends survive a successful delisting and need chasing separately. Google's autocomplete can keep offering the story the moment someone types your name, and images from the article can stay in image search after the web result has gone.
When Google says no
A refusal is common and it is not the end.
The refusals we see rest on four grounds, and each is answerable. The story is recent. You hold, or held, a role the story is relevant to. The offence is serious or the conviction is not yet spent. Or the article is court reporting, where Google gives weight to open justice. The answer is nearly always fresh evidence about what has changed since publication rather than further argument about the article.
Since 19 June 2026 there is a step that did not exist before. Section 164A of the Data Protection Act 2018, inserted by the Data (Use and Access) Act 2025, gives you a statutory right to complain to the controller directly. Google is required to acknowledge the complaint within 30 days and to investigate without undue delay. That gives a refusal a formal second look, and it builds a record that helps if the matter escalates.
Beyond that, you can complain to the ICO under section 165 of the Data Protection Act 2018. It costs nothing, and the ICO can decide that continued listing breaches data protection law. We do not always recommend that route, because in some cases an ICO complaint can cause more harm than good, and we advise on it case by case.
The court is the backstop. Section 167 of the Data Protection Act 2018 lets the court order a controller to comply, and NT1 and NT2 shows the claim can succeed. It is the last step rather than the first, and most matters never reach it.
Delisting cases we have acted in
Identifying details in each are changed for privacy.
- Deleting articles from Google after being found not guilty: an acquittal that kept following the client through search.
- Will Google delist articles about your professional life: the public-role ground, which is where Google resists hardest.
- Removing news articles from Google after dropped charges: charges that went nowhere, still first in the client's name results.
- Removing foreign conviction news articles from Google search: a conviction abroad, reported here.
- How doctors can remove outdated articles from Google: a regulated professional whose old story stayed at the top of the page.
Frequently asked questions
Will delisting remove the article from the internet?
No. The article stays on the website that published it, and anyone holding the direct address can still read it. What changes is that your name stops leading anyone to it. Removal at source is a separate application to the publisher.
Will Google tell the newspaper that I asked?
Google may tell a site that some of its addresses have stopped appearing for some searches, without naming you or identifying the search. In our experience publishers rarely react, and where a reaction looks like a realistic risk in your case we would tell you before anything goes in.
Can I apply to Google myself?
You can, and Google's form is open to anyone. The difficulty is that the grounds which succeed are not the ones people reach for instinctively, the application commits you to a factual position, and a refusal is harder to reopen than a first application is to get right. Where a case is simple enough to handle yourself, we will say so.
How many of these applications succeed?
Nobody can give you a reliable figure, and a solicitor quoting one should worry you. What we can do is read your case against the four factors above and tell you where it sits before you spend anything.
Does a delisting decision ever get reversed?
Google can reconsider if circumstances change materially, in either direction. A new story about the same subject can put a delisted URL back into play, which is one reason we look at the whole picture rather than the single link that prompted the call.
What if I want the article gone completely?
Then delisting is half the job and removal at source is the other half. Some publishers agree to remove or anonymise where the legal position is put to them properly, and where they refuse, the delisting still stands on its own.
If your name in Google still brings back a story you have moved on from, you do not have to accept it. Tell us what comes up and we will assess your legal position, free of charge and in confidence.
