Google unveils Gemini 4 Argon and gives it first to vetted cyber defenders, without cyber guardrails

Google on Wednesday announced Gemini 4 Argon, its next frontier model. In a post by Koray Kavukcuoglu, SVP at Google DeepMind and Google's chief AI architect, the company says the model performs at frontier level on complex, long-horizon workflows in software engineering, enterprise knowledge work such as legal and finance, and cybersecurity defense. Google says Argon has a 1 million token limit, and its own staff already use it for daily work such as debugging and codebase migrations. TechCrunch reports that Google also highlights its ability to analyze long videos and charts. The model is not broadly available yet. Google is first rolling it out to trusted cyber defenders through its Fairwind Program, a limited-access initiative for governments, Google Cloud customers and security partners that launched in early September with Gemini 3.8 Flash Cyber and the CodeMender harness. Google says it trained Argon specifically for defensive cyber work and that the model can autonomously find, validate and patch critical vulnerabilities. SecurityWeek reports that Google credits Argon with finding a critical flaw in software used by hospitals worldwide. For trusted defenders and its internal teams, Google says it will release Argon without cyber guardrails. The Verge reports that the limited launch is meant to let Google check that the model is not misaligned before a public release. Access will expand gradually as Google collects feedback from early testers.