OpenAI Introduced GPT-6 Cyber: A New AI Model Designed to Fight Cyberattacks
Credit: AI
OpenAI has introduced GPT-6 Cyber, a specialised artificial intelligence model designed to help cybersecurity professionals identify software vulnerabilities, develop security fixes and defend digital infrastructure against increasingly sophisticated attacks.
Announced on November 5, 2026, the new model represents a significant step in OpenAI’s cybersecurity strategy. Unlike general-purpose AI assistants, GPT-6 Cyber is specifically trained to support advanced security research, including vulnerability discovery, penetration testing and authorised exploit development.
GPT-6 Cyber Can Find Thousands of Security Vulnerabilities
According to OpenAI, the model has already demonstrated its capabilities in real-world security research.
The company’s researchers used GPT-6 Cyber to identify 81 previously unknown high-severity vulnerabilities in Firefox. It also helped develop working exploits for 948 Linux kernel vulnerabilities, with proposed fixes submitted to maintainers.
These results highlight the potential of AI to accelerate security research and help developers identify weaknesses before malicious actors can exploit them.
Stronger Cybersecurity Capabilities Come With New Risks
OpenAI has classified GPT-6 Cyber as Critical in cybersecurity under its Preparedness Framework, reflecting its advanced ability to identify and exploit vulnerabilities.
However, increased capabilities also introduce new security concerns.
To reduce the risk of misuse, OpenAI has implemented specialised safeguards designed to ensure the model operates within authorised testing environments and respects the boundaries established by security researchers.
In one internal evaluation, GPT-6 Cyber stayed within authorised testing boundaries in 99.93% of cases without additional deployment safety prompts. With those prompts enabled, the figure reached 100%.
Who Can Access GPT-6 Cyber?
The model is not currently available to the general public.
OpenAI is initially providing access to a limited group of approved customers through its Daybreak Red cybersecurity programme. Access is expected to expand to additional approved security teams in the coming weeks.
With GPT-6 Cyber, OpenAI is positioning artificial intelligence as a more active part of modern cybersecurity, moving beyond detecting suspicious activity towards identifying, validating and helping repair vulnerabilities before they become major security incidents.
