Google announces Gemini 4 and says it’s so capable that only ‘trusted cyber defenders’ can have it right now

Google has announced a new AI model called Gemini 4 Argon that the company says performs well on complex tasks in software engineering, enterprise knowledge work, and cybersecurity. The model is currently available only to a limited set of trusted cyber defenders while Google works with the U.S. government on its voluntary pre-release process and strengthens safeguards against misuse, prompt injection attacks, and misalignment before broader release. The announcement comes as Google is actively deploying the model in its own internal workflows and follows competing announcements from OpenAI about new AI models and agents.