Google opens Gemini 4 Argon first to cyber defenders
Google introduced Gemini 4 Argon with initial access for trusted cybersecurity partners in its Fairwind Program. The company is using the model internally for software work while gathering feedback and refining safeguards before broader access. Its announcement describes longer multi-step tasks, code migration and vulnerability repair, with developers, enterprises and consumers set to follow the early testing groups.
Artificial Intelligence··Morning
Cybersecurity partners receive early Argon access
Google introduced Gemini 4 Argon on September 30 and gave initial access to cybersecurity partners. The company is opening the model to trusted defenders through Fairwind, its program for early cybersecurity access, while its own teams also use Argon internally. Early testers are providing feedback while Google refines safeguards ahead of wider access for developers, enterprises and consumers. The company identifies paid API customers and Google AI Ultra subscribers as the starting groups for that later rollout, without setting a firm date.[1], [2]
Google uses agents for code migration
Google says its software teams use Argon agents to migrate C and C++ code into Rust. Work on the Zircon kernel, the core of Google’s Fuchsia operating system, involves more than 800,000 lines. Before critical changes enter production, they undergo automated and manual audits, emulation tests and review.[1]
Safeguards track reasoning and actions
Google’s safety work addresses misuse, indirect prompt injection, actions that depart from user intentions and isolation of high-risk test environments. It describes monitoring model reasoning and actions, with execution stopped where necessary. Wiz, a cybersecurity company, is using Argon in Scan for Good, a program that protects critical public infrastructure.[1]
