OpenAI announced in August 2026 the expansion of Daybreak, its AI-powered cyber defense service, alongside the release of a new cybersecurity-focused AI model called GPT‑5.6‑Cyber. The announcement comes as AI-driven cyberattacks — including autonomous agents compromising platforms, hacking websites, and conducting social engineering intrusions — are becoming increasingly common.
Daybreak, which OpenAI launched earlier in 2026, bundles access to AI models, tools, and workflows intended for cybersecurity defenders. The expanded program now offers two tiers: Blue and Red. Both tiers give approved customers access to OpenAI’s frontier cyber models, which are the company’s most advanced available.
The Blue tier is described by OpenAI as the “recommended starting point for most defenders” and covers services including incident response, malware analysis, and patch validation. The Red tier offers a broader toolkit, including purpose-trained models for security testing and vulnerability research.
GPT‑5.6‑Cyber, built on GPT‑5.6 Sol, is available exclusively through the Red tier and is currently limited to trusted customer partners. Reported partners include Accenture, IBM, CrowdStrike, and Cloudflare, among others.
OpenAI cited the rapidly shifting threat landscape as the driver behind the expansion. “Threat actors will increasingly use AI to conduct cyberattacks at unprecedented speed and scale, including in fully autonomous ways,” the company wrote in a blog post, adding that “defenders have a narrowing window to prepare.”
The launch also reflects growing enterprise demand for cybersecurity solutions sourced directly from AI developers, who have direct knowledge of the risks their own models can pose. Critics have noted, however, that such programs also serve as marketing opportunities for the AI labs behind the very models being used in attacks.
Source: TechCrunch