AI Text Watermarking Is Free And Good, Explained by Aaronson
TheZvi · x · 2026-08-22
Scott Aaronson largely solved AI text watermarking at OpenAI. The method uses a secret key to derive a pseudo-randomness source for token selection. By scoring how well text fits this pattern versus others, it reliably detects AI-generated content with minimal impact on quality.
Related event: AI Text Watermarking Sparks Debate Over Effectiveness and Model Rights(3 posts)→
More from Safety
- [un]prompted II AI security conference unveils agenda: 81 sessions over 3 days in SF — WeldPond · 2026-10-16
- Apate's AI agents get scammed on purpose, wasting 17 years of fraudsters' time — SumitGup · 2026-10-16
- Vector Institute partners with Town of Caledon in its first municipal AI adoption deal — VectorInst · 2026-10-16
- 'Most dangerous technology in the world' put in toddler sandboxes, and we're surprised when it escapes — KevinNaughtonJr · 2026-10-16
- NYC Comptroller urges Anthropic to pause IPO until AI safety and governance concerns are addressed — ShakeelHashim · 2026-10-16
- Dwarkesh: Anthropic insider expects AI to match humans at AI research within 9 months — ShakeelHashim · 2026-10-16