CybersecurityUpdated November 9

OpenAI limits GPT-6 Cyber to approved teams with access safeguards

At first, only selected, approved cybersecurity teams in Daybreak Red can access GPT-6 Cyber, through API and Codex. OpenAI lists IP restrictions and login monitoring among the safeguards, and says the model won’t help carry out attacks on critical infrastructure. The first rollout is deliberately limited both in who can use it and what it can help with.

1 publisher1 report

Coverage started November 9