Anthropic’s Watermarking, How It (Probably) Works, Worse Than It Seems

Anthropic is adding watermarking to its AI models in response to the European Union's AI law, a move that Ben Thompson of Stratechery criticizes as a terrible idea primarily for philosophical reasons. The watermarking technique, which is likely to be imperfect and potentially harmful, raises concerns about its effectiveness and implications.

Anthropic’s Watermarking, How It Probably Works, Worse Than It Seems By Ben ThompsonSource: Stratechery https://stratechery.com Anthropic /glossary/anthropic is adding watermarking in response to the E.U.'s AI law. It's a terrible idea, first and foremost for philosophical reasons. Get AI news in your inbox Daily digest of what matters in AI.