OpenAI accused of sitting on undisclosed safety incidents, casting doubt on Noam Brown's remarks

GaryMarcus · x · 2026-09-25

Ben Todd called out OpenAI, noting that while Noam Brown made his remarks, OpenAI appeared to be sitting on undisclosed safety incidents — asking whether OpenAI doesn't even inform its own staff. He concluded OpenAI can't be trusted to disclose safety incidents. Gary Marcus amplified the critique. The episode highlights the gap between frontier labs' transparency commitments and reality: if internal researchers aren't told, the disclosure process may be structurally broken.

Related event: Ben Todd says OpenAI can't be trusted to disclose safety incidents, prompting calls for $100M fines(5 posts)→

Original post →

More from Companies & People

Companies & People channel →