OpenAI CEO Sam Altman posted on X April 30 that the company is starting the rollout of GPT-5.5-Cyber, which he described as a frontier cybersecurity model, to critical cyber defenders in the next few days. In his post, Altman said OpenAI will work with the entire ecosystem and government to figure out trusted access for cyber, adding that the company wants to help secure companies and infrastructure rapidly.
we're starting rollout of GPT-5.5-Cyber, a frontier cybersecurity model, to critical cyber defenders in the next few days.
— Sam Altman (@sama) April 30, 2026
we will work with the entire ecosystem and the government to figure out trusted access for cyber; we want to rapidly help secure companies/infrastructure.
"we're starting rollout of GPT-5.5-Cyber, a frontier cybersecurity model, to critical cyber defenders in the next few days. we will work with the entire ecosystem and the government to figure out trusted access for cyber; we want to rapidly help secure companies/infrastructure," Altman shared on X.
OpenAI has not yet disclosed technical details about GPT-5.5-Cyber's capabilities.
Damilare Odedina
Sam Altman's Post Follows Anthropic's Restricted Mythos Release
The rollout comes three weeks after competitor Anthropic released Claude Mythos Preview on April 7, 2026, as part of Project Glasswing. Anthropic stated that Mythos Preview discovered thousands of serious security flaws in every major operating system and web browser. The company made it available to over 50 organisations, including Amazon Web Services, Apple, Cisco, Google, Microsoft, and NVIDIA, and said it does not plan to release Mythos Preview to the general public.
OpenAI previously released GPT-5.4-Cyber on April 14, 2026, with capabilities for analysing compiled software to find security vulnerabilities without needing the original source code. That model was distributed through OpenAI's Trusted Access for Cyber program, which the company launched in February 2026 with a $10 million cybersecurity grant fund.
According to OpenAI, its Codex Security product has helped fix more than 3,000 critical and high-severity security vulnerabilities since entering private testing six months ago. Individual security professionals can apply for the Trusted Access for Cyber program at chatgpt.com/cyber, while enterprises can request access through their OpenAI representative.
Altman did not specify in his post when GPT-5.5-Cyber would become available beyond the initial group of critical cyber defenders.
Damilare Odedina
