Crypto HeadlinesOpenAI implements invisible watermarking for ChatGPT texts
OpenAI is implementing the textGrain technology to watermark texts created by neural networks. The tool is designed for ChatGPT and Codex users in the European Union.
How it works
The watermark is integrated directly into the text structure using statistical patterns of word and token selection. The method does not use hidden characters or special spaces.
A specialized detector allows determining whether a text was fully generated or processed by OpenAI models.
Method limitations
- Detection accuracy decreases when paraphrasing text
- Translation into other languages affects analysis results
- Replacing synonyms can significantly reduce detector efficiency
The initiative is linked to compliance with EU AI Act requirements for mandatory labeling of content created by artificial intelligence.
19.3K views
