Anthropic Adds Watermarks to Claude AI-Generated Content

TL;DR. Anthropic now embeds imperceptible machine-readable watermarks in text and images produced by its Claude AI models. - This move aligns with the EU AI Act's transparency requirements for AI-generated content. - The watermarks help platforms identify Claude-produced content, boosting AI safety and accountability. - All new Claude models will feature these content marks from their initial release.

• Anthropic is implementing machine-readable watermarks for text and images generated by its Claude models.
• The watermarks are imperceptible to humans but allow identification of Claude-produced content by machines and platforms.
• This initiative supports compliance with the EU AI Act's transparency mandates for AI companies.
• New Claude models will incorporate these watermarking capabilities from their launch date.

Back to QLANKR News