Start with what the mark actually is, because it is not a stamp on the page. When a model writes, it picks each word from a set of reasonable options.
A watermark quietly splits those options into two groups using a secret key, then nudges the model towards one group. One word proves nothing. Across 500 or 1,000 words, a detector holding the key can see the pattern.
Anthropic began marking Claude output on 2 August, worldwide rather than only in Europe. It has not published the algorithm or released a detector.
It marks what Claude touched, not what Claude wrote
This is the part most coverage skipped, and Anthropic states it plainly in its own support article.
A detected mark means the content “may have been processed by Claude”. It does not mean Claude wrote it. Ask the model to proofread, translate or summarise your own writing, and the output carries the mark anyway.
The reverse also holds. No mark does not mean no AI. Short passages, heavy paraphrasing, older models and stripped file metadata all produce clean text that a machine still helped make.
So the signal has two failure directions at once, and the company says so upfront.
The law exempted the grammar fix
Here is where it gets strange. The EU AI Act does not require marking when a system performs an assistive function for standard editing, and the Commission’s own example of that is grammar correction.
Anthropic marks it regardless. Ars Technica called the approach “nuke it from orbit”.
The reason is structural rather than ideological. A watermark applied at the model level cannot distinguish a full draft from a comma. It marks the output, and the output is all it sees.
The result is a mark landing on exactly the content the law was written to leave alone.
And the label almost nobody has to show
Now look at the other half of Article 50, the part that governs what publishers must tell readers. It is far narrower than most people assume.
An AI-written novel needs no label. AI-generated marketing copy needs no label. Text that informs the public on matters of public interest does need one, unless a named and accountable human editor has reviewed it.
Put the two halves together and you get the contradiction. Every keystroke gets watermarked at the model level, while a fully synthetic article can reach readers unlabelled because an editor looked at it.
The EU has been building this machinery for a while. It has already made compulsory labels for synthetic content, and it has given itself powers to inspect and fine models directly. The penalty here runs to €15m or 3% of worldwide annual turnover.
The objections arrived within a day
Investor Bill Gurley argued that if only Anthropic can read the mark, it becomes “judge, jury, and prosecutor”. Anthropic told Business Insider it will ship a free detection API so anyone can check.
Former Microsoft executive Steven Sinofsky raised a different worry. The issue, he wrote, is “data retention and your right to private thoughts free of a digital trail”.
Simon Smith, who runs generative AI at the health agency Klick, asked whether a grammar check would now be flagged as AI-authored. Anthropic’s answer is the one above: the mark shows processing, not authorship.
The software trainer John Crickett asked a sharper question about code. If AI-generated code carries a mark, does that complicate a copyright claim in which the author must show human input?
The case for doing it anyway
The objections are not the whole story. Developer Donn Felker pointed out that marking helps models avoid training on their own output, the problem he called a “snake eating itself”.
Aadit Sheth of The Narrative Company made the reader’s case. Audiences should be able to tell whether the words they are reading reflect a person’s thinking.
The Commission’s own reasoning is broader still. It wants people to calibrate their trust, and it names fraud, impersonation and consumer deception as the risks.
Anthropic frames this as compliance rather than a stance. It is adding marking “to comply with the EU AI Act, and other labs are taking similar steps”, the company said. It also says the watermark does not change the meaning, quality or readability of Claude’s responses.
Some questions remain open. Ars Technica asked the company for a timeline on detection, for any testing on false positives and false negatives, and for how the marks square with the standard-editing exemption.
Anthropic sent a statement that did not address those points. The false-positive rate is the number that matters most here, because it decides how much weight a teacher or an editor should give the mark.
It will not stop anyone who is trying
The uncomfortable part is that the mark catches the honest and misses the deliberate. Paste watermarked text into a second model and ask for a rewrite, and the signal is likely gone.
File-based provenance is weaker still. A screenshot removes it, and so does any competent metadata editor.
The desk has seen this failure before. Meta’s own AI detector missed cropped images, which is roughly the amount of effort involved.
There is also a quiet cost. To keep the statistical signal intact, the model sometimes takes a slightly worse word than the one it would otherwise have chosen.
Anthropic is the only one doing text
This is the commercial risk, and it is real. OpenAI declined to comment specifically to TechRadar and pointed to a support page instead.
That page confirms OpenAI marks images and audio using SynthID and C2PA metadata. On text, it says expanding provenance to all modalities is a goal, as standards and tooling “continue to mature”.
Google uses SynthID too. Neither company has shipped text marking, and both face the same requirement to sell generative systems in the EU.
Existing models have until 2 December to comply. Until then, one lab is carrying the friction alone.
What the law does not touch at all
Article 50 requires disclosure that you are speaking to a machine. It says nothing about what happens to the conversation afterwards.
Stanford researchers reviewed the privacy policies of six major American AI developers. Every one used customer chat data to train by default, and some retained it indefinitely.
So the law now marks the output and ignores the input.
What would settle it
Two things, and both are checkable. The first is the detection API, and specifically its false-positive rate, because a mark on a student’s proofread essay carries the same weight as one on a fabricated news story.
The second is whether OpenAI or Google ship text marking before 2 December, or discover that the standards are still maturing.
Labels already misfire without any of this. X put a “Made with AI” tag on the resignation post of the departing White House press secretary on Wednesday, and then the tag quietly disappeared.
