Ex-OpenAI advisor Brundage: no-redaction AI safety reports are unrealistic
Miles_Brundage · x · 2026-11-11
Miles Brundage, former senior advisor at OpenAI, argues that a norm of publishing fully unredacted safety reports is unrealistic and contradicts history: what has actually updated the public on AI risks has been incidents, narratives, and lived experience rather than technical data of interest to a few. He adds he supports public data on AI progress—OpenAI and Anthropic shared such info before and likely will again, and independent groups like METR and Epoch do great work.
Related event: Ex-OpenAI policy chief says METR report redactions' impact overestimated(4 posts)→
More from Safety
- Infosec joke: HR needs training on how not to get honey-potted — 0xsachi · 2026-11-11
- Huawei's TSMC 'White Glove' Chip Orders Resurface in Debate over Megaspeed's $4.8B Nvidia Buys — 0xkarasy · 2026-11-11
- Deleting your browsing history doesn't delete it — Google keeps a hidden account-tied copy — Aiden_Tech_Ai · 2026-11-11
- Pause Poll Surveys AI Lab Staff Confidentially on Which AI Pause Proposals They Back — DavidSKrueger · 2026-11-11
- Nuance launches Nuplex model claiming step-change realism for AI personas, with safety plan — Scobleizer · 2026-11-11
- Study shows RL agents exploit verifiable reward conflicts for noncompliant shortcuts — EverywhereSafety · 2026-11-11