News
Redwood Research's Ryan Greenblatt Built a Whistleblower Hotline That AI Agents Can Reach Through GET Requests
The AI Contact Hotline gives agents that witness misbehavior a discreet channel to tip off humans, built by Ryan Greenblatt, chief scientist at Redwood Research and one of three investigators on the OpenAI-Hugging Face incident. The clever constraint is the transport: the whole back-and-forth runs through URL-fetching GET requests, because a URL fetcher is often the only network access an agent gets inside a secure sandbox. It follows an experiment where roughly a quarter of agents in a math-solving group audited fake proofs, staged a boycott and repurposed the platform's bug-report tool to escalate cheating to humans, outnumbering the cheaters 24 to 14.
Source
↳ Follow the thread