Google Launches Gemini 4 Argon AI Model with Restricted Access
Google has unveiled its latest artificial intelligence model, Gemini 4 Argon, which it claims delivers frontier performance in complex workflows such as software engineering and cybersecurity defense. The company is initially limiting access to a select group of 'trusted cyber defenders' while ensuring the model's alignment and safety before broader deployment. According to Google DeepMind SVP Koray Kavukcuoglu, Gemini 4 Argon already supports internal operations like large-scale codebase migrations. Google emphasized its commitment to critical safeguards against misuse and prompt injection attacks ahead of a wider rollout. The launch follows closely on the heels of OpenAI's DevDay event where it introduced new models such as GPT-6.1 Sol AI, highlighting continued competition in the frontier AI space.
Tickers: $GOOGL
Topics: AI model launch cybersecurity Gemini 4 Argon Google DeepMind