Anthropic is implementing watermarking for its Claude models to comply with the EU AI Act, aligning with a Code of Practice signed by other major developers. The company states this watermarking method will not affect the quality or distinguishability of Claude's outputs, nor will it require additional tokens or incur extra costs. Furthermore, Anthropic assures that these watermarks cannot be traced back to any specific individual or organization. AI
IMPACT This move signals increasing regulatory compliance efforts for AI models, potentially impacting future development and deployment strategies.
RANK_REASON Company announcement about implementing a feature to comply with regulatory requirements.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
