Anthropic Watermarks Claude, Pressuring AI Peers on Provenance

Anthropic’s decision to watermark Claude’s text outputs using a version of Google DeepMind’s open-source SynthID-Text system establishes a powerful de facto standard for AI content provenance. Announced on Friday to align with emerging European AI regulations, this move strategically escalates pressure on all major model providers to adopt similar traceability measures. Rather than developing a proprietary method, Anthropic