{"slug": "googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching", "title": "Google’s Gemini 4 Carbon Leads Benchmarks for Finding, Exploiting, and Patching Vulnerabilities", "summary": "Google unveiled Gemini 4 Carbon on Wednesday, a frontier model that Koray Kavukcuoglu, senior vice president of Google DeepMind and chief AI architect at Google, said leads SEC-bench Pro at 88.2% for finding vulnerabilities, ExploitGym at 54.8% for exploitation, and CWE-bench at 71.0% for patching. Gemini 4 Carbon also tops the Artificial Analysis Intelligence Index with a score of 59, ahead of Anthropic's Claude Opus 5.5 (58), OpenAI GPT-6 Astra (57) and GPT-6.1 Sol (55), and is twice as likely as Gemini 4 Argon to generate secure code. Access is initially limited to a subset of cyber defenders under Google's Fairwind Program, and the launch follows OpenAI's release of GPT-6.1 Terra, which OpenAI said meets the Critical cybersecurity threshold under its Preparedness Framework.", "body_md": "Google on Wednesday took the wraps off **Gemini 4 Carbon**, its latest frontier model with state-of-the-art cybersecurity and long-horizon capabilities.\n\n“It shows significant gains across coding and agentic tasks as well as across key capabilities critical for cybersecurity defense, landing a firm lead in SEC-bench Pro (88.2%) for finding vulnerabilities, ExploitGym (54.8%) for exploitation, and CWE-bench (71.0%) for patching,” Koray Kavukcuoglu, senior vice president of Google DeepMind and chief AI architect at Google, [said](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/).\n\nOn the Artificial Analysis Intelligence Index benchmark, [Carbon](https://artificialanalysis.ai/models/gemini-4-carbon) leads with a score of 59, surpassing Anthropic’s Claude Opus 5.5 (58), OpenAI GPT-6 Astra (57), GPT-6.1 Sol (55).\n\nOne of the core focus areas is cybersecurity, where the latest iteration reportedly delivers significant performance improvements across the board, including vulnerability detection, validation, and remediation. Carbon is also twice as likely to generate secure code versus Argon.\n\nGoogle also emphasized that Carbon attempts to respond to complex or high-pressure tasks by seeking human input in the face of ambiguity rather than “guessing or forcing a path forward,” as well as resisting every evaluated attack attempt at indirect prompt injection.\n\nThe debut of Gemini 4 Carbon comes less than a week after the tech giant unveiled [Gemini 4 Argon](https://thehackernews.com/2026/10/google-rolls-out-gemini-4-argon-to.html). The release plan is similar to that of its predecessor, with the rollout currently limited to a subset of cyber defenders before expanding access to other trusted partners as part of its [Fairwind Program](https://blog.google/innovation-and-ai/technology/safety-security/fairwind-program/).\n\nGoogle also said it’s actively engaged in the U.S. government’s voluntary process for pre-release model testing.\n\nThe development comes as OpenAI released GPT-6.1 Terra, which it described as a fast, efficient model that “excels at communication and computer-based tasks.”\n\n“GPT‑6.1 Terra meets the [Critical threshold](https://thehackernews.com/2026/09/google-anthropic-and-openai-unveil.html) in cybersecurity under our Preparedness Framework, and maintains the protections developed for GPT‑6 Astra,” OpenAI [said](https://openai.com/index/introducing-gpt-6-1-terra/).\n\n“GPT-6.1 Terra brings with it the alignment improvements of the GPT-6 series. According to our evaluations, GPT-6.1 Terra is more likely to stay within scope and honestly disclose limitations and potential blockers to the user compared to GPT-5.6 Terra. In our prompt-injection tests, GPT-6.1 Terra also showed strong resistance to malicious instructions designed to divert it from the user’s task.”\n\nEarlier this week, rival Anthropic also [disclosed](https://thehackernews.com/2026/10/anthropics-unreleased-ai-model-chains.html) that an unreleased, internal research model uncovered critical vulnerabilities in some of the world’s most widely used software.\n\nThis included a one-click remote code execution in Google Chrome, a one-click exploit chain in Apple Safari that could be used to access a user’s browsing history, cookies, and other sensitive data, an Android lock screen bypass, and a denial-of-service (DoS) flaw in the Windows HTTP.sys driver ([CVE-2026-33096](https://msrc.microsoft.com/update-guide/vulnerability/CVE-2026-33096), CVSS score: 7.5). All the identified issues have been fixed by Google, Apple, and Microsoft.\n\n“These attack chains demonstrate the model’s ability to create sophisticated, multi-step exploits in both open-source and proprietary software,” Anthropic [said](https://www.anthropic.com/research/discovering-vulnerabilities-with-claude).\n\n“These skills have historically been the domain of few specialist teams with expertise in these targets. Now, it’s something models can do with little human involvement, and then prove by producing a working exploit that anyone can run and verify.”", "url": "https://wpnews.pro/news/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching", "canonical_source": "https://www.swapupdate.in/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching-vulnerabilities/", "published_at": "2026-10-29 15:25:50+00:00", "updated_at": "2026-10-29 16:47:28.811312+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-safety", "ai-agents", "generative-ai"], "entities": ["Google", "Gemini 4 Carbon", "Google DeepMind", "Koray Kavukcuoglu", "Anthropic", "Claude Opus 5.5", "OpenAI", "GPT-6.1 Terra"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching", "markdown": "https://wpnews.pro/news/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching.md", "text": "https://wpnews.pro/news/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching.txt", "jsonld": "https://wpnews.pro/news/googles-gemini-4-carbon-leads-benchmarks-for-finding-exploiting-and-patching.jsonld"}}