OpenAI on Monday expanded its Daybreak program with two access tiers for cybersecurity defenders and launched GPT-5.6-Cyber, a specialized model for authorized vulnerability research.
Daybreak Blue offers general purpose models with defense-tailored safeguards, while Daybreak Red provides purpose-trained models including GPT-5.6-Cyber, the Microsoft-backed said.
OpenAI said it used GPT-5.6-Cyber to uncover two previously unknown Chrome V8 vulnerabilities, reported to Google and assigned as CVE-2026-15903, along with issues in a mobile operating system, a database, and an operating system kernel.
The company will require hardware security keys for all Daybreak accounts starting Sept. 1.
OpenAI said GPT-5.6-Cyber was assessed as High for cybersecurity capability under its Preparedness Framework, below the Critical threshold.
Price: $505.99, Change: $+6.00, Percent Change: +1.20%