Key facts
- Anthropic is implementing invisible watermarks in its Claude AI models.
- The watermarking is a measure to comply with the EU AI Act.
- Some users are protesting the implementation of AI watermarks.
- Concerns exist about the watermarks' impact on academic and professional use.
- Other users support the watermarking for transparency.
- The EU AI Act aims to regulate artificial intelligence.
Anthropic is introducing invisible watermarks into the text generated by its Claude AI models as a measure to comply with the European Union's AI Act. This new feature is intended to help identify content produced by artificial intelligence. The implementation has generated significant user backlash, with a segment of the user base expressing concerns about potential negative consequences for academic and professional applications of the AI. These users fear that the watermarks could expose their use of AI in contexts where it might be prohibited or discouraged. Conversely, another portion of the user community supports the watermarking initiative, viewing it as a positive step towards greater transparency in AI-generated content. The EU AI Act represents a comprehensive regulatory framework for artificial intelligence within the European Union, and Anthropic's decision to watermark its AI output is a direct effort to align its services with these new legal requirements. The specific details of how these watermarks will function and their detectability remain key points of discussion among users and potentially regulators.