SITUATION EXPLAINED: OpenAI is slowing the release of Astra over cyber capabilities. • OpenAI says it cannot rule out that Astra reaches critical cyber capability under its preparedness framework • Critical is the threshold at which a model is considered unsafe to release publicly, one step above High • It will scale up testing and security around the model and slow development until safeguards are in place • Anthropic made a similar commitment to pause training, then rolled it back in its…

MTS@MTSliveAug 7

SITUATION UPDATE: OpenAI has officially slowed development of its highly anticipated Astra model, telling Axios it cannot rule out that the AI possesses 'critical' cyberattack capabilities after running internal evals.