Anthropic has started inserting invisible code into Claude’s text to mark it as AI-generated.
The regulatory requirement
The move satisfies the EU AI Act’s Transparency Code. Regulators now demand labels for content that is AI-generated or edited in a way computers can identify. European officials are pleased with the compliance, but users are not.
Reddit shows the growing anger. Posts vary in tone. One user named visionode posted a rant about three weeks after creating the account. The post described the watermarking system as a draconian conspiracy designed to victimise innocent users.
The student and journalist
Visionode argued that savvy users can hide their usage by paraphrasing outputs through other services. The average user, however, will be caught.
“Who will get caught? You. The student who used Claude to reorganize a paragraph. The journalist who asked the AI to summarize a two-hundred-page transcript. The writer who had creative block and asked for synonyms. Those guys come out of the process with a digital tattoo on their forehead.”
Those examples are not the strongest. A journalist summarising a two-hundred-page transcript is not bothered by a watermark unless they copy and paste the summary verbatim into an article. That is plainly unethical and should not happen.
It is equally unethical for a student to copy and paste Claude’s output into an essay after asking it to reorganise a paragraph.
Community pushback
Other Redditors were not supportive of the outrage. One poster merely commented to get a load of the guy. Another asked the original poster to take a deep breath.
Visionode was not the only one complaining. Another user called the watermarks unethical and disgusting. They argued that by using Claude, they did the lion’s share of the work. In their view, the chatbot was merely a tool that facilitated their arduous labor.
“I gave the instructions, context, decisions, and countless refinements, claude was the tool. If Claude starts watermarking the code or anything else it generates, what exactly is it claiming credit for?”
Other users dogpiled onto the critic. One user shot back that the system is about detecting AI-generated outputs because of the risks they can cause in various situations. Another quipped that the critic could not even complain about Claude without using it to write the complaint.
Hypocrisy and utility
Some critics avoided the victimhood narrative to make slightly more nuanced arguments against Anthropic’s new policy. One poster complained of general hypocrisy in watermarking an editorial product that was itself generated by hoovering up other people’s work.
“I think it’s a very sinister direction to take,” said the user. “I don’t use Claude to write anything but having an AI that watermarks your work is terrifyingly ironic given how many of the frontier models came by their training data.”
In general, however, users have tended to support the watermarking system as a sensible way to track material generated by an algorithm. One user on another thread said there is literally no good argument for why this is not a good idea. The only reason not to want this is to lie to people.
What it means
Professionals who copy and paste AI text directly into reports or essays will face detection. The system targets verbatim output rather than the ideas or effort behind the work. Users who refine prompts and edit the results manually are less likely to trigger the flags.
