OpenAI releases ChatGPT 5.6 Cyber, but it's only for approved users

OpenAI has developed a new model called "GPT 5.6 Cyber," designed for vulnerability research, penetration testing, incident response, and remediation.

The incident highlights how adversaries continue to evolve their tradecraft, combining increasingly accessible tooling with targeted social engineering to slip past traditional perimeter defenses.

Finally, maintain offline, tested backups and a clear communication plan so that business continuity decisions are made ahead of time rather than under pressure.