Anthropic has recently decided to watermark Claude’s outputs, a move that has sparked mixed reactions. While some users are embracing the policy as a necessary step towards transparency, others are voicing concerns about being unfairly caught out.

The backlash began on Reddit, where one user, visionode, argued that the new system is a conspiracy designed to victimize innocent chatbot users worldwide. The example given was of students, journalists and writers who may inadvertently fall into the watermark trap by using Claude for simple tasks such as reorganizing paragraphs or generating synonyms.

However, others argue that these examples are too simplistic. For instance, summarizing a two-hundred-page transcript is unlikely to be an issue unless the user directly copies the summary verbatim—ethically dubious even without watermarking.The debate has also touched on the ethical implications of the AI itself. One user complained about the hypocrisy in watermarking content that was generated from other sources, questioning the irony of an AI marking its own work.

Despite the negativity, the majority support the watermarking system as a sensible way to track material generated by algorithms. Many see it as a necessary step towards accountability and transparency in a world increasingly reliant on AI.