TrellisMark: AI Text Watermarking That Tracks Users Across a Billion-Address Space
xeophon · x · 2026-08-24
TrellisMark is an AI text watermarking method that efficiently searches a billion-address space to recover user-specific identifiers from marked text. It demonstrates that building large-scale attribution systems is feasible using existing watermarking literature. The author warns against deployment but aims to make risks concrete for countermeasure research.
More from Safety
- Europe and America's Divergent AI Strategies: Regulation vs. Speed to Counter China — VraserX · 2026-08-24
- Is Anthropic irresponsible for keeping short timeline evidence private? — YafahEdelman · 2026-08-24
- UChicago Sosc Core to ban AI and enforce technology-free classrooms — RexDouglass · 2026-08-24
- LLM diagnostic accuracy drops 60% when communicating with patients — peterwildeford · 2026-08-24
- Secret Injection Cheatsheet: 6 Rules to Prevent AI Agent Leaks — blaizedsouza · 2026-08-24
- AI use implodes big book deals, forcing publishing industry reckoning — TuhinChakr · 2026-08-24