Anthropic to watermark AI-generated text to meet EU rules
Anthropic says its models, including Claude, will now watermark generated text and files to comply with the EU AI Act's Transparency Code. The feature applies automatically to models released after August 2 and will later extend to older ones.
Anthropic has confirmed it will watermark text and files produced by its AI models, including Claude, in order to comply with European Union regulations. The company detailed the change on an updated support page.
The EU AI Act's Transparency Code, which took effect on August 2, requires AI companies to mark AI-generated or edited content so that other systems can identify it. Anthropic said all models released after that date will automatically include technology to watermark both generated text and files. For files, the company relies on the open C2PA standard.
Watermark travels with copied text
Anthropic said it plans to extend watermarking support to its older models as well. Because the watermark is embedded within the text itself, it will travel along when the text is copied and pasted elsewhere, and may persist through some editing. The watermarking is applied at the model level, meaning it will be present regardless of which Claude product or surface generated the text.
It remains unclear how much editing would be required to strip the watermark; TechCrunch has asked Anthropic for clarification and is awaiting a response. The watermarking will apply across several of the company's products, including the Claude platform API, Claude, Claude Code, Claude Cowork and Claude Tag.
Industry response to backlash
Several platforms have recently moved to label AI-generated content, driven both by user backlash and a desire to avoid regulatory scrutiny. Last week, AI music platform Suno said it would mark tracks created through its service following a series of legal challenges. Last month, newsletter service Substack partnered with Pangram to flag AI-generated content, after its CEO raised concerns about a practice dubbed "Claudefishing."
Beyond Anthropic, other companies including Black Forest Labs, Google, Meta, Microsoft, OpenAI and Synthesia have also committed to following the EU's code.
