As we prepare to release Astra, we’re focused on making increasingly capable AI…
- Source
- x.com
- Date
As we prepare to release Astra, we’re focused on making increasingly capable AI safe and broadly accessible. Astra represents a significant advance in cybersecurity capability, reaching the Critical threshold under our Preparedness Framework. We're previewing how we evaluated the model, how its safeguards have advanced alongside its capabilities, and what we'll continue to learn and improve. https://t.co/OrrTgdU90K
Context
OpenAI's Astra safeguards include a monitor that can stop potentially unauthorized activity. OpenAI says it may also flag legitimate work, including tasks unrelated to cybersecurity or agents running for a long time.
When the monitor pauses a task, ChatGPT or Codex users may be asked to review the action; on other surfaces such as the API, the task stops. OpenAI plans to keep calibrating the safeguards to reduce unnecessary interruptions.
Checking sign-in…
Loading comments…