How Claude's text watermarking works \ Anthropic
Future Claude models will generate text that contains a watermark. This is a way of determining the likelihood that Claude was involved in writing the text, and we, along with several other major AI providers, are implementing this change to comply with the EU AI Act. In this article, we share answers to some of the questions we’ve received about how our chosen watermarking method works, whether it affects Claude’s outputs, and why we’re making this change.
Future Claude models will generate text that contains a watermark. This is a way of determining the likelihood that Claude was involved in writing the text, and we, along with several other major AI providers, are implementing this change to comply with the EU AI Act. In this article, we share answers to some of the questions we’ve received about how our chosen watermarking method works, whether it affects Claude’s outputs, and why we’re making this change. To summarize: We use a method of watermarking that does not have any practical impact on the quality or content of Claude’s outputs;…
saved by
related reading
- kat (@transkatgirl) on Xx.com
- Tenobrus (@tenobrus) on Xx.com
- The Mark of the Machinenewyorker.com
- How Claude marks AI-generated contentsupport.claude.com
- A global workspace in language models \ Anthropicanthropic.com
- Tracing the thoughts of a large language model \ Anthropicanthropic.com
- [2301.10226] A Watermark for Large Language Modelsarxiv.org
- Natural Language Autoencoders \ Anthropicanthropic.com
- Claude 4.5 Opus' Soul Document — LessWronglesswrong.com
- I can never talk to an AI anonymously againtheargumentmag.com
- Claude’s Constitution \ Anthropicanthropic.com
- Detecting AI fingerprints: A guide to watermarking and beyond | Brookingsbrookings.edu