OpenAI is adding text watermarking in ChatGPT and Codex

What happened: OpenAI is rolling out textGrain, an invisible machine-readable watermark in text output, to ChatGPT and Codex users in the European Union only, across all plans over the coming weeks. API customers worldwide can opt in starting today. Key numbers: No concrete figures provided. Why it matters: OpenAI says it is not making watermarking a global default at launch, citing room to learn from real-world use. It says textGrain matched or exceeded Google DeepMind's SynthID, and Anthropic made a similar move for the EU AI Act.

OpenAI is rolling out textGrain, an invisible machine-readable watermark in text output, to ChatGPT and Codex users in the European Union only, across all plans over the coming weeks. API customers worldwide can opt in starting today.

OpenAI says it is not making watermarking a global default at launch, citing room to learn from real-world use. It says textGrain matched or exceeded Google DeepMind's SynthID, and Anthropic made a similar move for the EU AI Act.