Anthropic is adding watermarking in response to the E.U.’s AI law. It’s a terrible idea, first and foremost for philosophical reasons.
Anthropic is adding watermarks to its AI models in response to EU AI Act compliance, but the approach has fundamental flaws in both concept and execution. Watermarking creates a false governance precedent—reducing transparency on model provenance while giving regulators and users false confidence in content attribution, a poor foundation for responsible AI stack decisions.
Anthropic is adding watermarking in response to the E.U.’s AI law. It’s a terrible idea, first and foremost for philosophical reasons.
Already subscribed?
$15 / month or $150 / year
With Stratechery Plus you get access to the subscriber-only Stratechery Update and Stratechery Interviews, and the Sharp Tech, Sharp China, Dithering, Greatest of All Talk, and Asianometry podcasts.
Stratechery UpdateSubstantial analysis of the news of the day delivered via three weekly emails or podcasts.
Anthropic defends and explains its AI watermarking feature
Anthropic’s EU Code Commitment Brings Watermarking Questions Into Focus
Has Anthropic's new watermarking feature opened up another can of AI worms? | The National
Tech analyst Ben Thompson dismisses the 'clearly absurd' concept embedded in AI watermarking
Explaining Anthropic’s New Watermarking Of Claude AI-Generated Outputs And What It Signifies For Society
Claude users are canceling their subscriptions, citing Anthropic’s new AI watermark
Claude Users Can’t Opt Out Of New Watermarks—Here’s What We Know
Anthropic to start watermarking Claude-generated text, images - SiliconANGLE
Some Claude users are mad that Anthropic's new watermarks will catch them cheating at their jobs, classes | TechCrunch
Anthropic says it will watermark text generated by its AI models | TechCrunch
