EU AI Act Watermark Verification Study
Watermarks Without Verification: AI Text Watermarking After the EU AI Act
Watermark claims from major providers remain unverifiable, creating governance gaps despite EU regulations requiring detectable AI content marking.
On 2 Sep 2026, Article 50 of the EU AI Act took effect, requiring generative AI providers to mark AI-generated content.
Anthropic disclosed that all Claude models released after that date embed SynthID-Text watermarks by default with no opt-out.
Google deployed SynthID-Text in Gemini since 2024.The study evaluated open-source SynthID-Text on two open-weight models, finding detection remains near chance.
Watermarks Without Verification: AI Text Watermarking After the EU AI Act
On August 2, 2026, the obligations of Article 50 of the EU AI Act took effect, requiring generative AI providers to mark the content their systems produce and ensure it can be detected as AI-generated. Days later, Anthropic disclosed that every Claude model released after that date embeds a watermark based on SynthID-Text in all generated text, enabled by default with no user opt-out; Google has deployed SynthID-Text in Gemini since 2024. Users objected that the watermark degrades quality, particularly for code, that it secretly encodes identifying information, and, in mutual contradiction, that it is easily removable and inescapable; the vendor answered with assurances of unchanged quality, no identifying information, and robustness to light editing. In this work, we argue that neither the objections nor the assurances can currently be verified and that this unverifiability, rather than watermarking itself, is the substantive governance failure. We sort the contested assertions by what it would take to settle each and evaluate the open-source SynthID-Text implementation on two open-weight models, because no public tool can test the deployed systems. On prose, the measured effect of the watermark does not exceed that of changing the sampling seed. On code, the cost is three points of correctness on one model and below measurement on the other, while detection remains near chance, a limitation of detectability rather than quality. The remaining gaps trace to withheld access or missing institutions and we map each to a requirement: release of matched outputs, configuration disclosure, accredited audits, a shared evaluation protocol, and interoperable detection.
This report is categorized under AITOP's "Papers" desk, focusing on Europe's AI governance and technology ecosystem. Within our cross-regional analytical framework, it directly maps to AI regulation, compliance frameworks, and sovereign compute policy and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include SynthID-Text, Claude, Gemini, EU AI Act.
Primary source verification: published directly by arXiv: Anthropic with first-party authority. Cross-source corroboration: 11 independent sources have covered this development (including shao__meng, Paul Graham, The Business Times: Tech), confirming industry-wide signal strength.
