On AI Risk
“It’s a cross-partisan position within misalignment teams across all frontier AI companies that business-as-usual AI development poses unacceptable catastrophic risk.”

Micah Carroll, Misalignment Preparedness, OpenAI · September 10, 2026

Title and organization are as of the date said.

Primary source verifiedExpertise: machine learning, AI safetyExtinction riskRace dynamics

The speaker’s own caveats

But we should also not hyperstition catastrophic risks into existence – they can be greatly reduced via safety requirements with teeth, international coordination, and a consensus to not build ASI unless there are sufficient safety advances to make us collectively confident to do so.

Source

Cite this quote

Carroll, M. (2026, September 10). It’s a cross-partisan position within misalignment teams across all frontier AI companies that business-as-usual AI development poses unacceptable catastrophic risk [Post]. X. https://x.com/MicahCarroll/status/2097865929959072069

More from Micah Carroll

All 2 quotes by Micah Carroll →

Others on the same topic

All quotes on extinction risk →All quotes on race dynamics →

Report an error in this quote