OpenAI launches Daybreak cybersecurity initiative — Codex Security takes on Claude Mythos

OpenAI unveiled Daybreak, a cybersecurity initiative that puts its Codex Security agentic system at the center of vulnerability detection and patch validation, paired with a network of partners across enterprise security teams and government-linked defenders. The launch is explicitly positioned against Anthropic's Claude Mythos, which last week disclosed 271 valid bugs across major OS and browsers. The cyber-AI arms race between the two labs is now the dominant frame for purpose-built defensive tooling.