Daybreak: Tools for securing every organization in the world

2026-06-22 · Source: OpenAI News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Software Development & Engineering · Depth: Advanced, long

Summary

OpenAI launched Daybreak on June 22, 2026, an initiative expanding tools and partnerships, including the full release of GPT-5.5-Cyber, to accelerate end-to-end patch automation for vulnerable software. Daybreak introduces an updated Codex Security plugin, which has already scanned over 30 million commits across 30,000 codebases, automatically fixing more than 500,000 findings. The full GPT-5.5-Cyber model, designed for authorized cybersecurity work, achieves a new state-of-the-art 85.6% on CyberGym, outperforming GPT-5.5's 81.8%, and also shows significant gains on ExploitGym (39.5% vs. 25.95%) and SEC-bench Pro (69.8% vs. 63.1%). Additionally, the Daybreak Cyber Partner Program allows security providers like Accenture and IBM to integrate GPT-5.5, while the "Patch the Planet" initiative, in collaboration with Trail of Bits and HackerOne, supports open-source projects such as cURL and Python in remediating vulnerabilities with expert human review. OpenAI is also collaborating with governments globally to protect critical infrastructure.

Key takeaway

For AI Security Engineers struggling with the volume of newly discovered vulnerabilities, Daybreak offers a critical shift towards automated patch generation and deployment. You should evaluate integrating the Codex Security plugin into your development pipelines to automate vulnerability identification, validation, and patch creation. Consider utilizing GPT-5.5-Cyber for advanced, authorized remediation tasks, enabling your team to move beyond just finding flaws to rapidly fixing them and enhancing overall cyber resilience.

Key insights

AI has shifted the cybersecurity bottleneck from vulnerability discovery to patching, necessitating automated remediation at machine speed.

Principles

• AI models excel at navigating codebases to identify and reason through attack paths.
• Effective cybersecurity requires validating, patching, and deploying fixes, not just finding vulnerabilities.
• Democratized access to advanced defensive AI capabilities is crucial for widespread protection.

Method

Codex Security operates by integrating into development workflows to identify plausible vulnerabilities, determine reachability, gather validation evidence, develop targeted patches, and verify results under human control.

In practice

• Deploy the Codex Security plugin for automated vulnerability scanning, reporting, and patch generation within developer workflows.
• Utilize GPT-5.5-Cyber for advanced, authorized cybersecurity tasks requiring high capability and permissiveness.
• Engage in programs like "Patch the Planet" to support open-source project vulnerability remediation.

Topics

• AI in Cybersecurity
• Vulnerability Management
• Patch Automation
• GPT-5.5-Cyber
• Codex Security
• Open-Source Security

Best for: CTO, AI Engineer, AI Product Manager, AI Security Engineer, Software Engineer, Director of AI/ML

Related on AIssential

• OpenAI says new GPT-5.5-Cyber outperforms Anthropic's Mythos on cybersecurity benchmark — OpenAI's Daybreak initiative shifts cybersecurity focus from vulnerability discovery to automated patching and remediation.
• OpenAI just released its answer to Claude Mythos — OpenAI's Daybreak uses AI to proactively identify and patch software vulnerabilities.
• Patch the Planet: a Daybreak initiative to support open source maintainers — AI-assisted security research, combined with human expertise, significantly accelerates vulnerability discovery and patching in critical open-source software.
• Google Cloud responds to AI-accelerated cyberattacks with a platform that aims to close security gaps in minutes — AI-powered cyberattacks demand automated defense platforms that deliver direct, verified security fixes, not just alerts.
• OpenAI Is Sending Security Engineers to Fix the Internet’s Most Neglected Code — AI can effectively augment human security engineers to proactively identify and fix critical open-source vulnerabilities.
• OpenAI Debuts Daybreak for Built-In AI Cyber Defence — OpenAI's Daybreak embeds frontier AI models directly into software development for proactive, continuous cyber defense.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by OpenAI News.