OpenAI is adding text watermarking in ChatGPT and Codex
Posts from this topic will be added to your daily email digest and your homepage feed.
TL;DR
- OpenAI is rolling out text watermarking to ChatGPT and Codex, starting with users in the European Union.
- The watermarking is an invisible, machine-readable feature designed to identify AI-generated text.
- OpenAI claims its textGrain watermarking performs comparably to other approaches but notes it doesn't guarantee detection or verify accuracy, ownership, or human contribution.
- API customers globally can opt in to use watermarked text outputs for select models starting immediately.
- Researchers and expert organizations can apply for access to the watermarking detection tools.