The EU’s transparency rules for AI-generated content apply from August 2, 2026, and Anthropic now marks what Claude writes and makes. The Claude watermark is a clue that a machine was in the room. It isn’t a record of who is accountable for what’s on the page, and a firm that treats it as one has handed its policy to a probability score.
Write the rule yourself. Who may use AI, who signs off, what clients are told. Then use the mark as a clue that backs it up.
Key points
- What does the Claude watermark mark? It marks text produced by supported Claude models and files such as images.
- Why can’t a watermark prove who wrote a document? Because the support page lists the ways it fails.
- Who carries the EU marking duty? Providers of generative AI systems carry it, not the businesses that use them.
- What belongs in your own policy and contracts? Put the answers in writing, with names on them, and put four questions to your vendors before a client puts them to you.
What does the Claude watermark mark?
It marks text produced by supported Claude models and files such as images. According to Anthropic’s support page, text carries an imperceptible watermark that travels with copied text, while files such as PNG and JPEG get signed C2PA Content Credentials, an open provenance standard that records who or what produced a file in signed metadata.
The page ties the marks to the EU AI Act’s transparency rules.
Anthropic says Claude models launched in the EU on or after August 2, 2026, support marking at launch, and that it is adding watermarks to outputs from earlier models, with the rollout still under way.
The wording is careful. Anthropic says a detected mark means content “may have been generated or processed by Claude”, which is a long way from written by Claude. Proofreading, translating and summarizing all count as processing. A passage that a person drafted and Claude only tidied could carry a mark. A passage Claude wrote and a person rewrote heavily might carry none. A signature in ink tells you who stands behind a page. A watermark tells you, roughly, that a machine touched it.
Why can’t a watermark prove who wrote a document?
Because the support page lists the ways it fails. Anthropic says text that was heavily edited, paraphrased, translated or mixed with other writing may lose the signal, short passages give an unreliable reading, and file credentials disappear when metadata is stripped through conversion, re-saving or screenshots.
Text from models whose marking hasn’t been switched on yet won’t carry a mark at all.
Then count the other tools in your office. Staff might also use ChatGPT, Gemini, Copilot or something a vendor embedded in your accounting software. Anthropic’s mark says nothing about those. A clean result from a Claude check can’t be read as proof that no AI was used. It means one vendor’s mark wasn’t found.
Picture a 60-person consultancy sending a client a report. It’s our illustration, not a measurement. Three staff touched it. One drafted with Claude and edited lightly, one used a different assistant, one wrote everything by hand. A Claude check might flag the first section, miss the second entirely, and say nothing about the third. Three different truths, one tool, and no way to tell them apart from the result alone.
Notice who the mark serves. It helps Anthropic show regulators it complied, and it helps platforms sort content. It doesn’t help you answer the question a client or a court will ask, which is who is responsible for this sentence. Only a person can answer that, and only a record can prove it.
Consider what follows for a Canadian firm that sends work to clients. The client doesn’t care which vendor’s mark is on the file. The client cares if you can say who produced the analysis and who checked it. That is a question about your process, and a mark can contradict you or fail to support you, but it can’t answer for you.
Who carries the EU marking duty?
Providers of generative AI systems carry it, not the businesses that use them. The EU AI Act’s Article 50 requires providers to mark synthetic text, image, audio and video in a machine-readable, detectable way, and it applies from August 2, 2026.
That guide reports a provisional agreement that gives systems already on the market until December 2, 2026, to comply with the marking requirement.
Deployers have separate duties. A summary of the Commission’s materials says deployers must disclose deepfakes and certain AI-generated text published to inform the public, with an exemption where a person has reviewed the content and accepts editorial responsibility. That exemption is the part to notice. It rewards exactly what a watermark can’t show, which is a named human who checked the work.
As we noted in our Washington frontier AI review piece, most AI rules reach you through your vendors. A Canadian firm with no EU customers may owe nothing under these rules. A firm that sells into Europe, or publishes content there, should ask a lawyer where it sits. We aren’t giving legal advice.
What belongs in your own policy and contracts?
Put the answers in writing, with names on them, and put four questions to your vendors before a client puts them to you. Find out which marks each AI vendor applies, on which products and plans, and whether they’re on by default in your account. Find out if you get access to detection, and on what terms.
Anthropic describes text detection as a private preview for eligible organizations, so most small firms can’t check anything today. Confirm whether removing, stripping or altering a mark breaches your terms, and do it before someone in marketing runs images through a converter.
Then fix your own side. Add a line to client agreements saying whether AI may be used on the work, who reviews it and who is accountable for the result. Keep a simple internal log of AI use on client deliverables. A spreadsheet with date, tool, task and reviewer is enough, and it’s dull on purpose.
Our piece on the Supreme Court walking away from AI copyright explains why a paper trail of who made what matters. Our look at the Munich ruling on AI disclaimers shows a court treating disclosure as no shield. Our look at shadow AI in Canadian workplaces explains why you can’t assume staff use only the tools you approved. The courts won’t settle it soon either, since AI-generated work copyright now rests on your paper trail.
Marks do help, and the sceptic’s case deserves its due. Regulators, newsrooms and platforms need some signal, and an imperfect one beats none, particularly for images where C2PA credentials are an industry standard. For public-facing media that may well be right. Our objection is narrower. Anthropic has told you in writing that its score isn’t proof, so building a policy on it outsources a judgment call to a vendor that disclaims it.
We haven’t run Claude’s detection and can’t see inside the watermark, so every statement about its behaviour comes from Anthropic’s own pages. Anthropic’s text watermark explainer, published August 14, says detection returns a probability rather than a yes or no answer and doesn’t state a false positive rate. Without that number, nobody outside Anthropic can say how often a mark would wrongly accuse a human writer. We’d revise our view if detection became broad, free, cross-vendor and published with error rates. Anthropic updates the support page as the rollout proceeds, so check the current version before relying on any detail here.
Name the reviewer for each deliverable by role, not by hope. If nobody can say who checked a client document, the watermark won’t help you, and the log will at least show you the gap before a client does.
Start the log this month.
Frequently asked questions
Does the Claude watermark prove my staff used AI?
No. Anthropic says a detected mark suggests content may have been generated or processed by Claude and is not conclusive proof. It also says a missing mark does not prove human authorship.
Does my business have to mark AI content under the EU AI Act?
The marking duty in Article 50 falls on providers of generative AI systems. Deployers have separate disclosure duties for deepfakes and some published text. If you sell into the EU, ask a lawyer how those apply to you. OpenAI took a different route in Europe, and the ChatGPT watermark does not cover Canadian users yet.
What should a small business put in its AI policy about watermarks?
Name who may use AI on which work, who reviews it, and what you tell clients. Keep a simple log of AI use on deliverables, because a log records accountability and a watermark does not.
Archive entry dated August 12, 2026, written and fact-checked on October 8, 2026. Sources are linked on the claims they support.
