OpenAI expands cybersecurity push with new model and partners

OpenAI on Monday announced a broad expansion of its Daybreak cybersecurity initiative, releasing a full version of its specialized GPT-5.5-Cyber model, an updated Codex Security plugin, a new partner program for security vendors, and an open-source protection effort called "Patch the Planet."

Upgraded Model and New Tools

The upgraded GPT-5.5-Cyber model scored 85.6% on OpenAI's internal CyberGym benchmark, up from 81.8% for its predecessor GPT-5.5. On ExploitGym, the model achieved 39.5% compared to 25.95% for GPT-5.5, and on SEC-bench Pro it reached 69.8% versus 63.1%. OpenAI said the model is designed for verified defenders whose work requires advanced cyber capabilities paired with verification, monitoring, and scoped controls.openai+2

The updated Codex Security plugin is designed to detect, validate, and remediate software vulnerabilities at scale. First previewed in March through Bloomberg, Codex Security works by identifying flaws, proposing solutions, and fixing bugs, enabling developers to focus on higher-level tasks.Investing.com+1

Patch the Planet and Partner Program

OpenAI also introduced "Patch the Planet" in partnership with security research firm Trail of Bits. The initiative leverages GPT-5.5-Cyber to find critical bugs in open-source software, with initial results including hundreds of bugs discovered, 64 pull requests submitted, and 51 issues filed across 19 projects in the first week. The program also works alongside vulnerability management platforms like HackerOne to provide free security consulting to open-source maintainers.Security Arsenal+1

The new Daybreak Cyber Partner Program brings together security vendors including Check Point ↘ 2.89%, Tenable ↗ 0.14%, TrendAI, Darktrace, Accenture ↗ 3.50%, Cisco ↘ 0.98%, CrowdStrike ↘ 0.51%, IBM International Business Machines Corporation ↗ 1.19%, and Palo Alto Networks ↘ 1.79%. Participating partners can embed GPT-5.5 with Trusted Access for Cyber into their security products and services.Check Point Software+3

Competitive Context

Daybreak represents OpenAI's answer to Anthropic's Project Glasswing and its Mythos AI model, which debuted to limited preview earlier this year and has helped identify vulnerabilities across various industries. Unlike Mythos, Daybreak is publicly available, and companies can request security assessments directly from OpenAI. The company said it plans to expand the partner program to additional organizations in the coming months.MacRumors.com+2