Anthropic · OpenAI · EU AI Act · Microsoft · Meta · The Register
OpenAI releases weak sauce watermarking for AI text
Compiled by KHAO Editorial — aggregated from 1 source. See llms.txt for citation guidance.
◌ Single Source
Compliance, they've heard of it.
Key facts
- In tests on 400-token passages, swapping out 10 percent of the words with synonyms reduced watermark detection from 92 percent to 66 percent
- The target error rate is one percent, but in short passages of 200 words, the detector may only flag 80 percent of the watermarks, compared to 95 percent in a 400-word passage
- The EU AI Act requires generative AI service providers to make model output identifiable in a way that's readable by machines
- OpenAI is now offering API customers the option to watermark the output of certain models to comply with the letter of Europe's AI Act
Summary
OpenAI is now offering API customers the option to watermark the output of certain models to comply with the letter of Europe's AI Act. But unlike rival Anthropic, it isn't applying its watermarking technique to content generated outside of Europe, though API customers can choose to have the marks applied to content made anywhere. The EU AI Act requires generative AI service providers to make model output identifiable in a way that's readable by machines. Anthropic was the first major AI company to announce its approach to AI Act compliance, which involves the application of Google's SynthID-Text algorithm.