AI Transparency & AI Governance
Claude Invisibly Marks AI-Generated Text: Why the AI Act Deserves Our Thanks
Anthropic embeds machine-readable watermarks in Claude texts. The technique is reminiscent of the barely visible yellow dots produced by color laser printers. What matters, however, is why we’re even hearing about this.
The Geviertstrich can breathe a sigh of relief.
A whole set of superstitions has now developed around the detection of AI-generated texts. Four dashes, certain sentence rhythms, unusually correct phrasing, or suspiciously neatly structured lists of three are quickly deemed proof that a text must have been generated by artificial intelligence.
That's not very durable.
Anthropic therefore takes a technically much more interesting approach with Claude. Supported Claude models embed an imperceptible, machine-readable watermark in the generated text. Anthropic actually describes the process as Claude “weaving” the mark directly into the text .
So are there secret, spiderweb-like structures between our words after all?
At least in a figurative sense.
Invisible traces of origin are not a new invention
The principle of embedding information about a document's origin—information that is barely perceptible to humans—is much older than generative AI.
It has been documented since the 2000s that certain color laser printers and color copiers leave tiny yellow dots on every page. The Electronic Frontier Foundation was able to decipher a repeating pattern of 15 by 8 dots on devices from the Xerox DocuColor series.
The dot pattern included, among other things, the printer's serial number and the date and time of the printout. Under normal lighting, these dots were barely visible. However, when magnified or viewed under blue light, they became discernible and could then be analyzed.
A printed document thus contained an additional layer of information that most users were unaware of.
The parallel with Claude is obvious: The AI-generated text also contains a trace that remains invisible to the reader but is intended to be recognized by machines.
However, the analogy has clear limitations. With the decrypted printer patterns, it was possible to link them to a specific device and time of printing. Anthropic, on the other hand, describes the Claude watermark as an indication that content may have been processed by Claude. The current documentation does not mention any association with a specific user, account, or device.
How Claude Labels AI-Generated Content
Anthropic relies on two different techniques.
1. Embedded Watermarks for Text
According to Anthropic, when a supported Claude model generates text, an imperceptible watermark is embedded directly into the text. It is not intended to alter the meaning, quality, or readability of the text.
Since the markup is part of the text, it should be included when copying and pasting. According to Anthropic, it can also survive some edits. The markup is applied at the model level and is therefore not limited to the classic Claude interface.
Supported models are expected to use the watermarks in Claude, via the API, in Claude Code, Claude Cowork, and on various cloud platforms, among other places. New Claude models introduced in the EU since August 2, 2026, support watermarking from the start. Anthropic is still working on a retrofit for older models.
This also means that, at this time, one should not assume that every piece of text ever generated by Claude automatically contains a recognizable watermark.
2. Signed provenance metadata for files
For supported file formats such as SVG, PNG, and JPG, Claude adds digitally signed provenance metadata.
These follow the open C2PA standard. C2PA is designed to document information about the origin and processing of digital content. A valid signature can indicate that a file was processed by Claude. It is also intended to make it possible to determine whether the file has been modified subsequently.
This is a different approach than the text watermark. With text, the signal is embedded in the output itself. With files, a signed information layer is also included.
However, metadata is not indestructible. Format conversions, resaving, or screenshots can remove such information. This is another reason why Anthropic does not rely exclusively on traditional file metadata for text.
We know that the marker exists. But we don't yet know exactly how it works.
Anthropic has announced that it will provide users and third parties with tools and technical capabilities to detect the markings.
However, the detailed technical documentation has not yet been provided.
So we now understand the principle and the planned scope. However, we cannot yet independently verify or replicate the text watermark. Anthropic itself states that details about the detection mechanisms will not be published until future documentation is released.
That's an important point.
A provenance signal only realizes its full value when independent systems can reliably detect it and correctly interpret the results. Otherwise, a technically interesting feature remains, for the time being, merely a claim made by the provider.
The presence of a watermark does not prove that the work was created by AI
The greatest danger is inferring more from the label than it actually implies.
According to Anthropic, if a Claude tag is detected, it simply means that the content may have been processed by Claude.
The original text could still have been written by a human. For example, Claude might have simply:
- corrected,
- translated,
- In summary,
- abridged,
- reformatted
- or have converted it to another file format.
Therefore, the discovery of a note does not prove that Claude developed the original idea or wrote the entire text on his own.
Similarly, the absence of a mark does not prove human authorship.
The flag may be missing if the text was generated using an older model, heavily edited, translated, combined with other content, or significantly shortened. In the case of very short text passages, there may also be insufficient material to generate a reliable signal.
At best, the watermark answers the question:
Was Claude likely involved in processing this content?
It does not answer:
Who came up with the idea, who verified the statements, and who is taking responsibility?
This means that the watermark is an indicator of origin. It is not a lie detector, not proof of plagiarism, and certainly not a seal of quality.
We really should say thank you to the AI Act for once
The EU AI Act is often associated primarily with bureaucracy, documentation requirements, and additional obligations.
This criticism is not entirely unfounded. However, the current case shows that regulation can also provide very tangible benefits.
Article 50 of the AI Act has been in effect since August 2, 2026. Providers of generative AI systems must design their systems so that artificially generated or manipulated audio, image, video, and text content can be marked in a machine-readable format and identified as such.
The technical solutions implemented should, to the extent technically feasible, be effective, reliable, robust, and interoperable.
In addition, the EU has established a code of conduct on transparency for AI-generated content. Participation in this code is voluntary. The legal transparency requirements under Article 50 are not.
Anthropic has signed the code of conduct and has stated that it will implement machine-readable labels to promote transparency and comply with legal obligations.
That is why we must allow ourselves to ask an uncomfortable question:
If it weren't for the transparency requirement in the AI Act, would we even know that Anthropic weaves such structures into its texts?
There is no definitive answer to that question. One should not assume that Anthropic would have remained silent in the absence of regulation.
One thing is certain, however: Without Article 50, there would be no comparable legal pressure to systematically introduce such origin indicators, document how they work, and provide third parties with the means to identify them.
Regulation does not, in this context, ensure that AI-generated text will automatically be accurate or of high quality. However, it does ensure that providers must be at least somewhat more transparent about their technical methods.
For that, we really do have reason to thank the AI Act.
What Companies Should Take Away From This
Stylistic features are not an examination method
Quotation marks, certain phrasing, or perfectly structured lists are not reliable indicators of AI use.
Such characteristics can provide clues, but they inevitably lead to numerous misjudgments. Professional governance must not be based on whether a text sounds like it was generated by AI.
Origin and responsibility must be considered separately
A technical indicator may suggest that an AI system was involved. However, it cannot determine whether the content has been fact-checked, is legally permissible, or makes strategic sense.
This responsibility rests with the company and the individuals who approve the content.
Human testing must be substantial
For certain AI-generated or manipulated texts on matters of public interest, the AI Act requires operators to label them. An exception may apply if there is genuine human review or editorial oversight, along with associated responsibility.
According to the European Commission’s guidance, a simple spell-check or superficial formal review is not considered sufficient human verification.
Companies should therefore not only document which tool was used; they should also specify who reviews the content for accuracy and who bears final responsibility for its publication.
Source signals belong in content governance
For images, videos, documents, and—in the future—text as well, provenance information is increasingly becoming part of professional content workflows.
These include:
- the collection of relevant metadata,
- the verification of digital signatures,
- clear approval processes,
- Transparent editorial responsibilities
- as well as guidelines for the use of generative AI.
Anyone who uses AI in marketing, communications, documentation, or customer processes should not wait until a regulatory agency or a customer asks about it before addressing these questions.
Transparency is infrastructure
The discussion about AI-generated content should move away from supposedly typical punctuation and phrasing.
What is more important are reliable traceability signals, open standards, transparent processes, and clear accountability.
The Claude watermark is not yet a perfect method. Its technical details have not yet been fully disclosed. Markings can disappear, and their interpretation remains limited.
Nevertheless, this approach is far more credible than the claim that AI-generated texts can be reliably identified by the use of em dashes or particularly correct language.
The AI Act does not automatically make AI-generated content true, good, or responsible.
However, it forces providers to make something visible that might otherwise have remained invisible without regulation.
Claude can weave the technical details into a text.
Responsibility for its content remains with the individual.