# OpenAI rolls out GPT-5.5-Cyber in a limited preview.

*genai · news · 2026-05-08 · Firstpost*

## Key points

- OpenAI has launched GPT-5.5-Cyber, a cybersecurity-focused AI model, in a restricted preview.
- GPT-5.5-Cyber is designed to be more permissive for vetted cybersecurity workflows, easing some safety restrictions.
- Preview access to GPT-5.5-Cyber is tightly monitored and targeted at critical infrastructure defenders.
- OpenAI's move follows Anthropic's recent release of Claude Mythos Preview, also limited to select security teams.
- Major financial and government leaders are now directly engaging with AI firms about cybersecurity-focused AI models.

**Companies:** OpenAI, Anthropic
**Countries:** United States

[Read the full story on Firstpost](https://www.firstpost.com/tech/openai-introduces-gpt-5-5-cyber-sam-altman-stresses-the-need-to-secure-companies-quickly-14008810.html)

---

Canonical: https://newsio.io/n/cbffea76-1155-45f9-a192-f71623c60ad7/openai-rolls-out-gpt-5-5-cyber-in-a-limited-preview-unlike-a-traditional-next-ge
Summarized by Newsio from Firstpost. https://newsio.io/how-it-works