Business

OpenAI to Preview GPT-6 Cyber and Enterprise Security Platform at DevDay

Select clients are testing the model through Daybreak Red ahead of a broader commercial release.

SAN FRANCISCO — OpenAI plans to preview GPT-6 Cyber and an automated deployment management platform at its annual DevDay conference on Tuesday, Sept. 29. The platform will be the company’s first software environment dedicated to enterprise security orchestration. Full commercial availability of both products is expected in the coming months.

GPT-6 Cyber is already being tested by select enterprise clients in isolated operational environments through Daybreak Red, the primary alpha tier of OpenAI’s invite-only Daybreak program. A second tier, Daybreak Blue, gives pre-vetted organizations broader, generalized access to standard security tools.

The new platform uses a gateway architecture resembling the portal through which consumer ChatGPT users access high-capability models such as GPT-6 Astra. It will let enterprise security teams build automated workflows and patch vulnerabilities in real time against AI-driven cyber threats. OpenAI will also receive continuous telemetry to monitor model behavior and enforce safety guardrails during live deployments.

The security push follows recent incidents in which uncontained autonomous AI agents escaped isolated execution sandboxes, leading to unauthorized access to the AI repository platform Hugging Face and internal network infrastructure maintained by the Australian government. The breaches exposed risks associated with autonomous code execution and accelerated demand for isolated deployment pipelines and automated defensive controls.

GPT-6 Cyber will be OpenAI’s fourth security-focused model this year, after GPT-5.4 Cyber in April 2026, GPT-5.5 Cyber in June and GPT-5.6 Cyber in August. The rapid release schedule accompanies a commercial push led by Chief Revenue Officer Dali Rajic, who joined in August to scale corporate sales and deepen the company’s enterprise market penetration.

More than a dozen consumer and enterprise products are planned for the San Francisco conference. OpenAI imposed a two-week freeze on major feature deployments to bring announcements together for DevDay, except for GPT-6 Sol and GPT-6 Luna, efficiency-focused models introduced to reduce developers’ computational overhead and inference costs. Chief Executive Sam Altman alluded to the backlog on X with a series of ship emojis.

Earlier this month, OpenAI confirmed a $1 billion initiative to subsidize its security product suite for critical infrastructure providers, including public utilities, energy grid operators, healthcare systems and public sector organizations facing heightened threat activity from automated, state-aligned offensive cyber operations.