Known for
Led Anthropic's work proving which model answered you.
Why chosen
He connects cryptography with accountability for AI systems. A Harvard computer science PhD researcher, he works on verifying what AI systems actually do and was part of FAR.AI's 2026 AViD workshop on AI verification. On November 2, 2026 he was lead author of FlashProve, Anthropic's open method for proving that a response came from a specific model. Proof over promises.
Roy Rinberg personal site: Harvard CS PhD student working on security in machine learning
royrinberg.com · Roy Rinberg
Role
Researcher, Anthropic
Signal
BuilderLead author of FlashProve, Anthropic's provable-inference research published November 2, 2026.
Current location
New York
Personal site
royrinberg.com
On the list since
Last checked