OpenAI is adding text watermarking in ChatGPT and Codex

<figure> <img alt="Open AI logo on a graphic green backround." data-caption="" data-portal-copyright="Image: The Verge" data-has-syndication-rights="1" src="https://platform.theverge.com/wp-content/uploads/sites/2/2025/10/STK155_OPEN_AI_CVirginia__C.jpg?quality=90&#038;strip=all&#038;crop=0,0,100,100" /> <figcaption> </figcaption> </figure> <p class="wp-block-paragraph">An <a href="https://openai.com/index/eu-text-provenance/">invisible, machine-readable watermark</a> in text output is rolling out to ChatGPT and Codex, but only for users in the European Union at first. </p> <p class="wp-block-paragraph">OpenAI says its textGrain watermarking "matched or exceeded" other approaches like Google DeepMind's SynthID for text, which is also the basis for the watermarking Anthropic <a href="https://www.theverge.com/ai-artificial-intelligence/980869/anthropic-claude-watermarks-synthid-text-system">announced in August</a>. Like OpenAI, Anthropic made the move to meet the requirements of the <a href="https://www.theverge.com/ai-artificial-intelligence/974571/eu-ai-act-transparency-labels-rules-deepfakes">EU's AI Act</a>; however, not everyone was happy to learn about the addition.</p> <p class="wp-block-paragraph">OpenAI also included scores from AI benchmarks showing similar performance from watermarked and unwatermarked text. But it notes that textGrain "does not guarantee …</p> <p><a href="https://www.theverge.com/ai-artificial-intelligence/1004880/openai-chatgpt-text-watermarks-eu-ai-act">Read the full story at The Verge.</a></p>