Miles Brundage: Preventing AI Theft and Tampering Is the Top Policy Priority
Miles_Brundage · x · 2026-09-18
Former OpenAI policy lead Miles Brundage argues that security—defined as AI technology being impossible to steal or tamper with—should be the highest AI policy priority, and explains why he shifted focus to it.
- Timing disconnect: Independent experts, drawing on interviews with frontier-lab staff, believe such companies are years away from defending AI tech against well-resourced attacks by countries like China and Russia—even if they try hard.
- Capabilities are advancing fast: AI firms already have systems better at competitive coding than all but a few hundred humans, with much more progress expected soon; such tech being stolen and abused is a real risk.
- Definitions: security means not being stolen or tampered with; safety means absence of accidents—the two should not be conflated.
- He also points to a recent RAND report on protecting algorithmic secrets.
More from AGI Musings
- 1988 sci-fi short story imagined an emergent Chinese Room that answered back — toptickcrypto · 2026-09-18
- OpenAI's Noam Brown: Air-gapping may not stop misaligned AI, safety bar must rise — basedjensen · 2026-09-18
- Martin Casado endorses the take that 'AGI' and 'ASI' are thought-terminating clichés — seanmcdonaldxyz · 2026-09-18
- Martin Casado on Noam Brown's air-gap warning: covert channels are an old, understood problem — tszzl · 2026-09-18
- People now use AI to message their own friends, and it may seed a startup — mobileraj · 2026-09-18
- Dwarkesh: Labs Will Hide Models During RSI; Delangue Calls Concentration the Biggest AI Risk — soumitrashukla9 · 2026-09-18