Superpower Daily3 min read
Frontier Model Forum publishes shared framework to rate AI jailbreak risks
The preliminary method weighs what a safeguard bypass unlocks and how readily it can be used. Testing will examine whether different assessors reach consistent, meaningful ratings.
By Superpower DailySuperpower Daily

Read the full story
Originally published by Superpower Daily on Nov 7, 2026.
Comments
Sign in to join the discussion.