OpenAI announced an upgraded Daybreak network defense toolchain and the full release of GPT-5.5-Cyber for cybersecurity defense use cases.
According to Odaily, in the CyberGym benchmark that evaluates AI agents’ ability to reproduce known vulnerabilities, GPT-5.5-Cyber scored 85.6% as a single model, compared with 81.8% for GPT-5.5 and 83.8% for Anthropic Mythos 5.
OpenAI said that as AI improves vulnerability discovery efficiency, the main bottleneck in cybersecurity is shifting from finding vulnerabilities to automatically fixing them. The company also upgraded its Codex Security plugin to support automated vulnerability analysis and patch generation within the Codex environment.
OpenAI said that since the preview release in March this year, Codex Security has scanned more than 30 million code commits and automatically confirmed fixes for 500,000 security issues. The updated version also supports CodeQL query integration and SARIF-standard file exports.
Separately, OpenAI said it launched the open-source security initiative “Patch the Planet” with organizations including Trail of Bits and HackerOne. The project will provide ChatGPT Pro subscriptions and API credits to support more than 30 major open-source projects, including cURL and Go, with a security expert team manually validating patch quality.
For the Daybreak partner program, OpenAI said security vendors including Palo Alto Networks and Wiz have begun integrating the related capabilities.
AI TRENDS | OpenAI Upgrades Daybreak Toolchain and Releases GPT-5.5-Cyber Model
2026-06-23 03:13:49
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
Next article：
AI趋势 | OpenAI推出GPT-5.5-Cyber完整版
