OpenAI expands Daybreak with GPT-5.6-Cyber for 16 security partners

OpenAI expanded its Daybreak Cyber Partner Program on August 10, giving approved red team and security specialist partners access to its GPT-5.6-Cyber model for offensive and defensive security work. The program is structured so that access to the underlying model stays strictly with the partner—never transferred to the client. Partners choose between Daybreak Blue, which supports broad defensive workflows, and Daybreak Red, which covers specialized and tightly governed activities like red teaming and penetration testing, where teams deliberately break into a customer's systems to reveal how far a real intruder could get. Sixteen companies are named: nine security and services firms (Accenture, IBM, Capgemini, Cognizant, EY, KPMG, PwC, NCC Group, and SpecterOps) and seven technology partners (Palo Alto Networks' Unit 42, CrowdStrike, Cisco, Sophos, Akamai, Fortinet, and Cloudflare).

Depending on the engagement, partners can use the model for vulnerability discovery and validation, red teaming, penetration testing, incident response, and remediation across complex enterprise environments. The key value lies not just in finding flaws but in determining exploitability, assessing risk, and ensuring fixes reach production. Milan Patel, global head of MDR services at Sophos, described the arrangement as delivering “frontier-grade defense at scale to organizations unable to deploy these models on their own.” Safeguards vary by engagement and can include identity verification, defined testing scopes, logging, monitoring, and human oversight—there is no fixed control set. Partners and their customers define boundaries, review findings, and apply judgment before action, meaning the actual controls are a matter for each contract. This partner-only distribution strategy lets OpenAI put powerful cyber capabilities in the hands of trusted firms while avoiding direct model access for end clients.