Paraphrase Pass opens
First edition of the directory: 7 provenance entries, 11 detectors, 14 rewriting tools, 10 papers and benchmarks, 8 policy texts, 12 news sources. Every link was checked on this date.
OpenAI announces textGrain
OpenAI's text watermark goes opt-in for API customers worldwide, off by default, and begins rolling out to ChatGPT and Codex output in the EU. Detector access is by application only. OpenAI's own figures: about 95 percent detection on 400-token passages at a 1 percent false positive rate, falling to 17 percent after a quarter of the words are swapped for synonyms.
Anthropic announces a global Claude text watermark
A SynthID-Text variant applied to future Claude models worldwide, with earlier models to follow. A detection API enters private preview for regulators, media, researchers, educators and enterprises with EU compliance needs. Anthropic says light editing probably leaves the mark; a full rewrite removes it.
EU Article 50 and California provider duties apply
The AI Act's transparency obligations take effect for providers and deployers, with a transition period for systems already on the market. On the same date, covered providers under California's AI Transparency Act owe disclosures and detection tools.
Commission finds the transparency Code adequate
The European Commission's opinion confirms that the Code of Practice on Transparency of AI-Generated Content covers the Article 50 obligations, giving signatories a presumption of conformity.
Final Code of Practice published
Two layers of marking, secured metadata plus watermarks, and a commitment from signatories to keep marks as non-removable as they can.
California signs AB 853
The AI Transparency Act is amended: provider duties move to August 2, 2026; large online platforms and generative AI hosting platforms are covered from January 1, 2027, including a ban on knowingly stripping provenance data.
Turnitin adds AI bypasser detection
Text that humanizer tools may have rewritten is flagged and counted toward the AI-generated total. English only.
SynthID-Text published in Nature and open-sourced
Google DeepMind publishes the method behind the Gemini text watermark and releases a reference library with Hugging Face.
SynthID text watermarking ships in Gemini
Google extends SynthID from images and audio to text in the Gemini app and web experience.
Detectors shown to flag non-native English writers
Liang et al. report an average 61.3 percent false positive rate across seven detectors on essays by non-native speakers, the finding that still anchors the false-accusation debate.