Anthropic pledges to embed watermarks to help discern AI slop in sop to EU
Anthropic
Anthropic has pledged to embed watermarks in its AI-generated content to help identify it, in a concession to the European Union. The move is part of broader efforts to address concerns about AI-generated misinformation. The watermarking is expected to help users distinguish between human and AI-created text.
Anthropic, the AI company behind the Claude models, has committed to embedding watermarks in its AI-generated content. This pledge is seen as an effort to appease the European Union, which has been pushing for greater transparency and accountability in AI. The watermarks would help discern 'AI slop'—low-quality or misleading content produced by AI—from authentic human output. This initiative aligns with EU regulatory pressures and reflects a growing trend among AI developers to implement provenance measures. The exact technical details of the watermarking were not specified in the report.
Source: Anthropic (GNews) —
original