Codex Security Plugin scans, patches bugs

OpenAI's latest update turns Codex into an active defense tool, transitioning AI coding assistants from passive autocomplete engines to proactive security auditors.

• –Deep integration: Instead of just highlighting vulnerabilities, Codex now creates codebase-specific patches, dramatically cutting down remediation time.
• –Cybersecurity pivot: The expansion of the Daybreak initiative and cooperation with Trail of Bits shows OpenAI is taking a top-down approach to critical open-source infrastructure safety.
• –Competitive positioning: This update moves Codex directly into competition with specialized security auditors like Snyk, while competing with Anthropic's Claude Code on terminal agent capabilities.
• –Model upgrade: Behind the scenes, the release of GPT-5.5-Cyber to trusted defenders hints at the next generation of reasoning models tailored specifically for code analysis.