Users Backlash Against Anthropic's Watermarking Feature

Some users have posted critical opinions on social media regarding the new watermarking feature introduced by Anthropic. The underlying concern is that unauthorized use of Claude in workplaces and schools could be exposed through this functionality.

A watermark is identification information embedded in AI-generated text or images, enabling verification of whether output was created by AI. By incorporating this feature into Claude, Anthropic has made it possible to track the origin of user-generated content. This technology is spreading amid growing calls for transparency and authenticity assurance as AI-generated content becomes more prevalent.

Most criticism on social media comes from users who have been secretly using Claude in workplaces or schools. It reflects dissatisfaction over the emerging risk that reliance on AI—despite employers or educational institutions requesting submissions as "self-created work"—could be exposed. Conversely, such reactions demonstrate that the watermarking feature can have a certain deterrent effect.

Watermarking technology itself is not new, and other major AI companies like OpenAI and Google are also considering and implementing it. As AI-generated content becomes integrated into daily life, distinguishing whether content was actually created by humans has become increasingly difficult, and digital watermarks are positioned as one solution. Anthropic's move aligns with this broader industry trend.

This incident further highlights social tensions surrounding AI usage. There remains a difficult-to-bridge gap between users who want to actively leverage AI as a productivity tool and organizations and society seeking transparency and fairness. As watermarking technology becomes more widespread, concealing the fact that AI was used will become increasingly difficult.

The future focus lies in how reliably this watermark functions and how organizations and educational institutions utilize such information. As technical accuracy improves, organizational rules and ethical discussions regarding AI usage may advance to more concrete stages. For users, this could become an opportunity to reconsider where and how AI should be used.