Press Space to continue
OpenAI is using cautious language around a frontier model's
Press Space to continue
This would be the first model to ever hit 'critical' on OpenAl's cyber Preparedness Framework, but they don't actually ever say that it has. Here they say we're 'treating it as' our first and in the blog post they say 'we cannot rule out' Critical capability level at this time. https://x.com/AndrewCurran_/status/20858…
OpenAI@OpenAIAug 7
After evaluating one of our upcoming models, Astra, we're treating it as our first "critical" model for cybersecurity under our Preparedness Framework. This is a scenario we've planned for, and we're putting additional controls in place to ensure Astra's further development happens safely and securely. We're working hard to make Astra broadly available, and get its advanced cyber capabilities into the hands of defenders. https://openai.com/index/responding-next-frontier-critical-cyber-capabilities/
open on x · 1 viewViews and clicks count once per visitor per day. Clicks are opens of this post on X.