Google unveils Gemini 4 Argon, initially limited to trusted cyber defenders
Google introduced its Gemini 4 Argon AI model, citing superior performance in software engineering, enterprise tasks and cybersecurity, and said it will first be available to a select group of vetted cyber-defense teams.
Google announced its latest AI model, Gemini 4 Argon, describing it as delivering frontier performance in complex workflows such as software engineering, legal and finance knowledge work, and cybersecurity defense. The company said the model is already being used in internal processes, including large-scale codebase migrations, and that benchmark results show it outperforming competing systems from other AI firms.
Access to Gemini 4 Argon will initially be restricted to a "set of trusted cyber defenders," with Google participating in a voluntary U.S. government pre-release access program while gradually expanding availability. The firm said it will implement additional safeguards before a broader rollout, aiming to protect against misuse, prompt-injection attacks and model misalignment.
The rollout follows OpenAI's recent developer event, where it introduced new AI agents and models, and comes shortly after Koray Kavukcuoglu was appointed head of DeepMind.