Anthropic Details How Claude's Invisible Text Watermarks Will Work
In one line: Anthropic revealed how it plans to embed machine-detectable, human-invisible watermarks into text generated by Claude.
Key points
- Anthropic announced on Friday that Claude's text-marking system is "a version of the SynthID-Text approach."
- SynthID-Text is an open-source watermarking technology from Google DeepMind that creates detectable patterns by adjusting word-choice probabilities.
- The goal is compliance with Europe's AI transparency rules, which require labeling of AI-generated content.
Why it matters
Distinguishing AI-generated text sits at the center of debates over misinformation, plagiarism, and copyright. The move signals a broader push by major AI firms toward verifiable provenance markers under emerging regulation.