OpenAI launches GPT‑5.4‑Cyber for cybersecurity
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI announced GPT-5.4-Cyber, a fine‑tuned variant of GPT-5.4 built for defensive cybersecurity use cases.
- OpenAI is offering the model to a limited group of participants in its Trusted Access for Cyber program.
- OpenAI frames the release as a direct competitor to Anthropic’s Mythos model for cybersecurity.
- Bloomberg reported the rollout on April 14, 2026, and the story was covered by dozens of outlets including The New York Times, Infosecurity Magazine, The Information, and Reuters.
- OpenAI’s Trusted Access program restricts the model to a limited group of vetted participants, keeping it out of general public hands.
Why it matters: Security teams that join OpenAI’s Trusted Access program gain a specialized AI model for defensive cyber operations, while the limited rollout keeps the technology out of general hands and intensifies competition with Anthropic’s Mythos, reshaping how AI is deployed in cybersecurity.
