OpenAI Flags ‘High’ Cybersecurity Risks in Future AI
OpenAI warns upcoming AI models could pose “high” cybersecurity risks, from zero-day exploits to industrial intrusions, and unveils defenses like access controls, tiered access, and a Frontier Risk Council.
