Ex-Anthropic researcher denies OpenAI leak accusations as firing drama escalates
harris_edouard · x · 2026-10-24
Former Anthropic researcher Ryan Greenblatt publicly denied OpenAI's accusations that he facilitated architecture leaks to The Information, which three fired OpenAI researchers were told he did. He says a journalist did contact him before the article, but he only asked to clarify that a referenced blog post of his was preexisting and public. He also denies reading an internal OpenAI monitorability memo reportedly containing sensitive architecture info. Evan Hubinger, who managed him on the Alignment Faking project, backed him: Ryan handled sensitive info with utmost professionalism, and "I trust Ryan a lot more than OpenAI here."
Related event: Ex-OpenAI researcher Greenblatt denies leak accusations(2 posts)→
More from Companies & People
- A 2-person AI team built its own video tool and dropped Zoom entirely — danshipper · 2026-10-24
- Every's 'frontier team' tracks new AI models so the rest of the company can build — every · 2026-10-24
- Meta's new commercial terms put legal liability for AI mistakes on customers — esporx · 2026-10-24
- Cloudflare AI lead is still hiring 'cracked engineers' for Workers AI and AI Gateway teams — michellechen · 2026-10-24
- Googler laments Grace Hopper cancellation: billions for AI but not for women in tech — marilynika · 2026-10-24
- ARC Prize Research Summit kicks off at MIT in Boston with Chollet headlining — xennygrimmato_ · 2026-10-24