Tech

OpenAI Announces Their Text Watermarking Plans

Read full article at source ↗

Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 06/10/2026, 03:22:24 UTC.

Briefings

OpenAI announced today that it will implement text watermarking for select EU models via API starting October 5, 2026, to comply with the EU AI Act. The technology, called textGrain, adds an invisible statistical signal to model outputs. Access to the detector is initially restricted to approved researchers and expert organizations. OpenAI notes that while textGrain matched or exceeded other approaches in ideal conditions, strong performance does not guarantee reliable detection in everyday use. The company explicitly states it is not forcing this on users outside the EU and has not yet made its detector publicly accessible.

Key points

• OpenAI will implement textGrain watermarking for select EU models via API starting October 5, 2026, to comply with the EU AI Act.

Context and limitations

Coverage is limited to the EU market; access to the detector is restricted to approved researchers and expert organizations; performance in ideal conditions does not guarantee reliable detection in everyday use.

Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 06/10/2026, 03:23:22 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.

View archived briefing on IPFS
bafybeieo364lydk57axlkzd5pwkprd7xhwamfkuk7lg5hzpsyczeccluou