GPT-5.6-Cyber (9 min read)

OpenAI has expanded its Daybreak access framework for approved security professionals and released GPT-5.6-Cyber, a cybersecurity-specialized model. Daybreak provides advanced cybersecurity capabilities to help defenders prepare for AI-driven threats before attackers do.

Daybreak operates in two tiers.

• Daybreak Blue: Provides access to general-purpose frontier models such as GPT-5.6 Sol, applying safeguards tailored for vulnerability discovery, secure code review, malware analysis, incident response, and patch verification.
• Daybreak Red: Offers cybersecurity-specialized models for approved vulnerability research, exploit verification, and security testing.

GPT-5.6-Cyber is built on GPT-5.6 Sol and has been trained to improve performance on advanced security tasks such as zero-day vulnerability discovery and exploit chain development, while reducing refusals for dual-use requests. However, access is restricted to trusted defensive personnel.

In OpenAI's internal Advanced Cybersecurity Completion Rate evaluation, GPT-5.6-Cyber completed 95.0% of advanced requests, including exploit chain development, authentication bypass, and privilege escalation. GPT-5.6 Sol recorded 1.5% with system protections enabled, 2.0% when using Daybreak Blue, and the previous GPT-5.5-Cyber recorded 57.3%.

Additionally, in the ExploitGym2 evaluation, which measures the ability to convert known vulnerabilities into real exploits achieving arbitrary code execution in controlled environments, GPT-5.6-Cyber outperformed both GPT-5.6 Sol and GPT-5.5-Cyber.