Google unveils Gemini 4 Carbon AI model for cybersecurity, praised by Elon Musk
Google introduced Gemini 4 Carbon, an AI system aimed at detecting software vulnerabilities and generating secure code, and received a brief endorsement from Elon Musk on X.
Google announced its newest AI system, Gemini 4 Carbon, positioning it as a tool for cybersecurity defence and complex software engineering tasks. The model is described by the company as capable of identifying software flaws, testing exploitability, and assisting in patch development. In benchmark tests reported by Google, Gemini 4 Carbon achieved an 88.2% score on SEC-bench Pro, 54.8% on ExploitGym, and 71% on CWE-bench, and it generated secure code at more than twice the success rate of its predecessor, Gemini 4 Argon.
Elon Musk responded to the announcement on X with a single word, "Impressive," to which Google CEO Sundar Pichai replied with a folded-hands emoji. The brief exchange mirrors a similar reaction Musk gave to a prior post by Pichai.
Google said the model will initially be available to a limited group of cybersecurity defenders through its "Fairwind Program," with a gradual rollout to trusted partners as safeguards are refined. The company noted that the cautious approach reflects broader industry concerns about the dual-use nature of powerful AI tools.
Pricing for Gemini 4 Carbon is set at $2 per million input tokens and $10 per million output tokens, matching the launch rates of the earlier Gemini 4 Argon model.