OpenAI, Anthropic Roll Out AI-Generated Content Watermarks to Meet EU Compliance Deadline
Major AI labs deploy imperceptible watermarks on images, audio, and text as EU AI Act Article 50 compliance deadline looms on December 2.
AI-generated summary. This news item was written automatically by an AI pipeline from published reporting and was not reviewed by a person before publication. It may contain errors. Read the original source.
Watermarking Push Ahead of EU Compliance Deadline
OpenAI and Anthropic are rolling out AI-generated content watermarking systems as the EU AI Act’s Article 50 transparency rules take effect. The EU AI Act Article 50 transparency rules took effect on August 2, 2026, with existing systems given a transition window and a compliance deadline of December 2, 2026.
OpenAI’s Multi-Signal Approach
OpenAI’s images and audio will carry two signals: C2PA metadata and SynthID watermarks. SynthID is an imperceptible watermark developed by Google DeepMind, which OpenAI has licensed.
Developers building on OpenAI’s platform get access to a Content Provenance API for eligible use cases, expanding the company’s watermarking infrastructure beyond its own products.
Industry-Wide Adoption
Anthropc introduced imperceptible text watermarks in new Claude models starting August 2, 2026. Notably, Anthropic chose to mark Claude’s output worldwide rather than running one version for Europe and another for everywhere else.
This approach mirrors the strategy adopted by Google, which has offered SynthID-Text in Gemini since 2024.
Context: Earlier Development and Regulatory Backdrop
OpenAI’s internal watermarking prototype was reported to be about 99.9% effective in 2024, but the company held back a broad release at that time.
OpenAI’s signing of the EU Code of Practice extended promises it had already made under the General-Purpose AI Code of Practice.
Source: Crypto Briefing