OpenAI has developed GPT 5.6 Cyber, a model designed for security work including vulnerability research, penetration testing, and incident response, according to BleepingComputer. The rollout is not a broad consumer release: OpenAI says access is limited to approved companies and supported security vendors. BleepingComputer reports that OpenAI named Accenture, IBM, Capgemini, Cognizant, EY, KPMG, PwC, NCC Group, and SpecterOps among the select companies with access. The model is also rolling out to supported security vendors including Palo Alto Networks, CrowdStrike, Cisco, Sophos, Akamai, Fortinet, and Cloudflare. OpenAI says regular users will not receive direct access to the underlying models because of security risks, according to BleepingComputer. Instead, approved partners are expected to use the capabilities inside existing security products, managed services, and customer engagements. OpenAI is offering the cyber capabilities through what it calls Daybreak Access, with two versions described in the report. Daybreak Blue is intended for a broad range of defensive security workloads. Daybreak Red is meant for more specialized work under closer governance. Depending on the engagement, partners may use the models for vulnerability discovery and validation, red teaming, penetration testing, incident response, and remediation across enterprise environments, BleepingComputer reports. OpenAI says the controls may include identity verification, defined testing scopes, logging, monitoring, and human oversight. The company’s stated boundary is that access to the underlying models remains with the approved partner and is not transferred directly to the customer. Partners are expected to work with organizations on the scope of each engagement, review findings, and apply their own expertise before action is taken. OpenAI is also opening a Daybreak Cyber Partner program for cybersecurity providers and consultancies that want to apply, according to the report. Organizations interested in using the technology are directed toward participating security providers rather than direct access from OpenAI. Who benefits: Approved security vendors, consultancies, and managed-service providers gain a new capability to embed in customer engagements. Enterprises that already work with participating providers may get access without building specialized cyber AI infrastructure themselves. Who's exposed: Regular users and organizations outside the approved partner ecosystem do not get direct access to the underlying models. Customers will depend on participating providers’ scoping, oversight, and implementation practices.