Unverified50% confidenceFactExact time
OpenAI's text watermarking technology is called TagScreen; it leaves detector-identifiable patterns in word selection through fine-tuning the model, and replacing approximately 10% of words with synonyms causes the detection rate to drop from about 92% to 66%.
1
Sources
50%
Confidence
Long-term
Relevance
10/8/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAfter text is manually rewritten or polished, the accuracy of AI detection methods based on textual features drops significantly.76% similarUnverifiedTextGrain forms patterns recognizable by detection tools through fine-tuning word choices, but editing, translation, or the use of very short excerpts may weaken the signal.72% similarUnverifiedIndustry text watermarking schemes typically introduce specific statistical biases during the model's token generation process, causing the output text to carry patterns that are algorithmically detectable but imperceptible to humans.72% similarUnverified一个合格的AI文本水印技术需要满足可检测性、鲁棒性（文本被改写、翻译或删减后签名仍存在）和低误报率71% similarVerifiedThe weakness of character-level watermarking schemes is that simple text normalization, copying to a plain-text editor, or regex filtering can completely remove them.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/989104API
curl https://kongchang.com/api/v1/knowledge/claims/989104MCP
get_claim(id=989104)