AI WireTechCrunch
OpenAI to watermark ChatGPT text for EU users
OpenAI says it will add invisible watermarks to text generated by ChatGPT and Codex in the European Union, according to TechCrunch. The move is meant to meet requirements under the EU's AI Act, which pushes companies to label AI-generated content. The watermark is hidden inside the text itself, not visible to readers, letting tools later check if a passage came from these systems. OpenAI notes that editing the text afterward can weaken the mark, making it harder to detect. The step shows how AI companies are adjusting their products to follow new regional rules on transparency around AI-generated content.
Words to know
- Watermarking
- — Hidden markers added to AI text so it can later be identified as AI-generated.
- AI Act
- — European Union law setting rules for how AI systems, including content labeling, must operate.
- ChatGPT
- — OpenAI's AI chatbot that generates human-like text responses.
- Codex
- — OpenAI's AI system designed to generate and understand programming code.
Summary written from TechCrunch's headline and teaser