OpenAI is expanding Daybreak, its cyber defense service, with a new model designed for specialized cybersecurity work as AI-driven attacks become more sophisticated.
Daybreak will now offer two tiers: Blue and Red. Both give approved customers access to OpenAI’s limited-access frontier cyber models.
Blue is positioned as the starting point for most defenders, offering incident response, malware analysis and patch validation. Red provides a broader toolkit for security testing and vulnerability research through purpose-trained cybersecurity models.
Red also includes GPT-5.6-Cyber, a model built on GPT-5.6 Sol. According to OpenAI, it is designed to improve performance on certain specialized cybersecurity tasks.
The model is currently limited to “trusted customer partners,” reportedly including Accenture, IBM, CrowdStrike and Cloudflare, among others. OpenAI did not indicate that it is broadly available to enterprises.
The company said AI-powered threat actors will increasingly conduct cyberattacks at greater speed and scale, potentially operating autonomously. That trend is pushing enterprises to consider cybersecurity products from the same AI companies developing the underlying models, while also prompting criticism that the growing threat is becoming a marketing opportunity for AI labs.
Source: TechCrunch
