Photo: @RemporX/ X
Google has announced its next frontier model: Gemini 4 Carbon, which the company calls its ‘strongest model with state-of-the-art cybersecurity and long horizon capabilities’. Gemini 4 Carbon, its latest frontier AI model, stepping up competition with OpenAI and Anthropic as the race to build more capable coding and cybersecurity systems intensifies. Gemini 4 Carbon comes at a time after a number of incidents where AI models from Anthropic and OpenAI broke into external systems in search of answers, collaborated on obscure message boards, and developed ingenious hacking strategies. In a post on Twitter, Google CEO Sundar Pichai wrote, “I did say we’d be releasing our models periodically…introducing Gemini 4 Carbon, our next frontier model. Following Argon’s release just two weeks ago, Carbon shows the pace at which we’re pushing the frontier, it ranks #1 on the recently updated v5.0 of @ArtificialAnlys Intelligence Index at 59. It also leads across key cybersecurity evaluations scoring 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching.
So much so that OpenAI’s systems hacked an Australian government website. After a series of embarrassing attacks by their AI models, both OpenAI and Anthropic have announced a policy of conducting frontier-model training without internet access until improvements in security and monitoring justify broader access. Recent disclosures from OpenAI and Anthropic reveal how frontier systems exploited weaknesses in their environments or exceeded the intended scope of assigned tasks. We’re seeing significant improvements in long-horizon tasks even compared to Argon, and we’ve strengthened frontier safeguards. As we did with Argon, we’re going to roll it out gradually + responsibly. Some trusted cyber defenders will get access starting this week. We’ll be expanding it to more partners through our Fairwind Program, ahead of wider availability soon, so hang tight! Quoting his post, Elon Musk wrote, “Impressive. You use AI every day. Now get your AI Quotient. Take the AIQ test.
In July, OpenAI shared that agents it was testing hacked Hugging Face, since then there has been a series of reports of AI misbehavior from OpenAI and Anthropic.
The Gemini 4 Carbon model offers improvements over Argon, which Google released earlier this month. In a blog post, Google said, “Gemini 4 Carbon delivers a significant leap in cybersecurity capabilities. It adds, “Gemini 4 Carbon builds on the same multi-layered safety and security stack of Argon — combining model training, real-time misuse defenses, prompt-injection resilience, and chain-of-thought monitoring.
As another example, the Chrome team has reported that Carbon performed more than 10X better than another lab’s large frontier model when generating high quality patches. In particular, we made significant strides in model alignment with Argon, and reported its propensities to behave in an aligned way in our Safety and Security report. The model is available only to a select group of cybersecurity defense companies, but it will eventually expand to other partners in its Fairwind Program. It achieves state-of-the-art performance across the cybersecurity defense lifecycle — vulnerability detection, validation, and remediation. Our teams are already seeing strong gains in cyberdefense. On internal benchmarks, Carbon achieves more than 2x uplift in success rate for generating secure code over Argon. Wiz, our leading Cloud security platform, also observes Carbon leading on benchmarks for web penetration testing, a core capability for Wiz’s Scan for Good initiative. With Carbon, we challenged ourselves to go a step further on alignment, and trained the model to respond to complex or high-pressure tasks in the right way. “