Introducing GPT-Daybreak to accelerate defenders

GPT-5.6-Cyber completes 95% of advanced cybersecurity requests, up from 1.5% for GPT-5.6 Sol, and found CVE-2026-15903 in Chrome's V8.

• Daybreak Blue offers GPT-5.6 Sol with guardrails removed for defensive work; Daybreak Red offers purpose-trained cyber models.
• GPT-5.6-Cyber found CVE-2026-15903, a high-severity V8 vulnerability, plus 400+ kernel privilege-escalation flaws.
• Individual Daybreak accounts must adopt hardware security keys starting September 1, 2026.
• Access is available to approved individuals and organizations via openai.com/daybreak/partners.

Security researchers and defenders can now run authorized vulnerability research and exploit development with far fewer refusals, accelerating defensive work that earlier models blocked.

Sources

• OpenAI: Introducing GPT-Daybreak to accelerate defenders

Read this as text

Back to the AI news