AIToday

Claude Opus 4.7 can identify anonymous writers from unpublished text — ending practical anonymity for anyone with a large public writing record

Hacker NewsApr 25, 20262 min read
Claude Opus 4.7 can identify anonymous writers from unpublished text — ending practical anonymity for anyone with a large public writing record

Summaries like this, in your inbox every morning.

Sign up free →

3 Key Points

  1. Anthropic released Claude Opus 4.7, which correctly identified the author (Kelsey Piper) from multiple unpublished writing samples — including a school progress report, movie review, fantasy novel excerpt, and 15-year-old college essay — none of which matched the author's published work. ChatGPT and Gemini performed worse at the same task, making this an Opus 4.7-specific capability.

  2. The AI identifies writers by detecting invisible stylistic patterns in prose, not through account information or memory settings. Testing in Incognito Mode and through the API produced identical results, meaning the deanonymization happens purely from text analysis — the AI learns who you are within a few substantive exchanges.

  3. For anyone with substantial published writing under their real name, online anonymity is now functionally gone. Academics, researchers, journalists, and others who maintain anonymous accounts while publishing publicly face involuntary identification. People with little to no public writing remain safe; the threat only applies to those with a large existing corpus online.

Discussion

No discussion yet for this article

Stay ahead with AI news

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Get Started Free

Free · takes 30 seconds · unsubscribe anytime

1 minute a day. The AI essentials.

200+ sources · Email / LINE / Slack

Get it free →