OpenAI is introducing a new feature that will watermark the text generated by its ChatGPT and Codex models for users in the European Union. This initiative is part of the company’s efforts to adhere to the EU’s AI Act, which aims to regulate AI technologies and ensure transparency.

The watermark is designed to be invisible and machine-readable, allowing for the identification of AI-generated content. OpenAI claims that its textGrain watermarking technology is advanced, reportedly matching or exceeding the effectiveness of similar technologies developed by competitors like Google DeepMind.

This move highlights the growing importance of compliance with regulatory frameworks in the AI sector, as companies navigate the complexities of ensuring responsible AI usage while maintaining user trust.


Compiled automatically by the Tech AI Newsdesk from public AI-news sources and summarised in our own words.

Frequently asked questions

What is the purpose of the text watermarking feature?

The text watermarking feature aims to identify AI-generated content and ensure transparency as part of compliance with the EU's AI Act.

How does OpenAI's watermarking technology compare to competitors?

OpenAI's textGrain watermarking technology reportedly matches or exceeds the effectiveness of similar technologies developed by competitors like Google DeepMind.