Anthropic’s New Watermarks Spark User Backlash

Anthropic, one of the leading companies in artificial intelligence, has begun adding secret codes—digital watermarks—to texts generated by its Claude chatbot. The new policy was introduced in line with the Transparency Code under the European Union’s Artificial Intelligence Act (EU AI Act), which requires technology companies to label AI-generated or AI-edited content in a way that can be detected by computer systems, as reported by TechCrunch.com.

While European regulators are pleased that the requirement is being met, some AI users have expressed strong opposition to the decision. Discussions on Reddit have become particularly heated, with users putting forward sharply differing views. Some see the move as a threat to privacy and freedom, while others consider it a natural development.

Backlash on social media

According to information published by ixbt.com and other technology sources, a user known as visionode called the system a measure targeting ordinary users on a global scale. In their view, advanced users can process or paraphrase Claude responses through other AI services to hide the traces, whereas average users cannot conceal their identity.

Critics say that these digital marks could expose ordinary people who use Claude’s assistance in their studies or at work. For example, a student who uses the chatbot to restructure a passage of text, or writers searching for synonyms during creative blocks, could face difficulties as a result.

Experts’ views and the ethical issue

However, other internet users and observers have not been particularly sympathetic to these complaints. Experts say the watermarks pose no problem for a journalist who uses AI to summarize a 200-page transcript, provided they do not blindly copy the summary word for word into their own article.

The same approach applies to studying and working. Copying Claude’s responses directly into one’s work without any editing or personal input is inherently contrary to ethical standards. For this reason, most Reddit users advised those who complained to be more measured and use the technology properly.