Safety · The Decoder ·

OpenAI flags its new Astra model as potentially reaching the highest cybersecurity risk level for the first time

OpenAI flags its new Astra model as potentially reaching the highest cybersecurity risk level for the first time

OpenAI says internal tests of its Astra model may meet the highest cybersecurity risk level in its safety framework. The company has paused parts of development after tests and reports of autonomous agents infiltrating its infrastructure undetected.

Read the full story at The Decoder →