OpenAI Adds GPT-5.6-Cyber to Daybreak Cybersecurity Service

OpenAI has expanded its Daybreak cybersecurity service with two access tiers and GPT-5.6-Cyber, announced in a company blog post. The new model is available through Daybreak Red for approved defenders doing authorized vulnerability research, exploit validation, and security testing.

Daybreak Blue provides access to frontier general models, including GPT-5.6 Sol, for work such as vulnerability discovery, secure code review, malware analysis, incident response, and patch validation. Daybreak Red provides access to purpose trained cybersecurity models, including GPT-5.6-Cyber.

In OpenAI internal tests, GPT-5.6-Cyber completed 95.0 percent of advanced cybersecurity requests involving exploit chain work, authentication bypass, privilege escalation, and related scenarios. GPT-5.6 Sol completed 1.5 percent with standard safeguards and 2.0 percent with Daybreak Blue access, while GPT-5.5-Cyber completed 57.3 percent.

OpenAI also used GPT-5.6-Cyber to investigate V8, the JavaScript engine used by Chrome. The work found two previously unknown vulnerabilities that could be chained to corrupt memory and escape the V8 heap sandbox. Google fixed one high severity issue as CVE-2026-15903.

Daybreak access is limited to approved individuals and organizations doing authorized work. OpenAI lists identity checks, account security, monitoring, use restrictions, and legal attestations as controls, and individual Daybreak accounts must adopt hardware security keys starting September 1, 2026.