None
EN
OpenAI flags possible critical cybersecurity risk in upcoming model, tightens controls
['Var Template_Content', 'Sso_Login_Box', 'Xwelcome Backorenter The Email Address Or Mobile Number Associated With Your Account To Sign In. Show Passwordsign Innew To The Indian Express Signupxcreate Your Account It Is Quick', 'Easy.Or Show Passwordnextvalidate Otpregisteralready Have An Account', 'Signin', 'Sso_Social_Box', 'Sign In Withgmailfacebookapple', 'Var Follow_Widget_Data', 'Af_Article_Count', 'Ie_Mobile_Check']
The Indian Express
OpenAI CEO Sam Altman has said the company is working to make its increasingly powerful AI models broadly available while strengthening safeguards around their capabilities.
(Image: Reuters)OpenAI said on Friday it cannot rule out that its upcoming AI model, Astra, has “critical” cybersecurity capabilities, prompting the startup to pause some internal development and trigger safety protocols.
Under OpenAI’s safety guidelines, a model reaches the “critical” threshold if it can autonomously identify and exploit severe, real-world software vulnerabilities, known as zero-day exploits, or execute complex cyberattacks against highly secure targets without human intervention.
This follows an exclusive report by Reuters that OpenAI has discovered more instances in which autonomous agents have escaped containment as the company expands its investigation of the hacking incident at tech firm Hugging Face that drew global attention in July.