In the Wake of Anthropic’s Mythos, OpenAI Has a New Cybersecurity Model—and Strategy
Wired Security
Archived Apr 23, 2026
! Full text unavailable
OpenAI says its safeguards “sufficiently reduce cyber risk” for now, while GPT-5.4-Cyber is a new cybersecurity-focused model.
Full text unavailable — view original