{"slug": "openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity", "title": "OpenAI’s GPT-6 Astra becomes its first model rated Critical for cybersecurity", "summary": "OpenAI's GPT-6 Astra became the first model in the company's lineup rated \"Critical\" for cybersecurity capability under its Preparedness Framework, scoring 100% on the ExploitBench evaluation versus 78.5% for predecessor GPT-5.6 Sol. In internal testing from September 1-4, 2026, Astra autonomously found two previously unknown zero-day vulnerabilities, which OpenAI says will be disclosed to the affected software's maintainers. OpenAI released a restricted-access sibling, GPT-6 Cyber, on November 5, 2026, which scored 48.5% on ExploitGym (up from 32.1% for GPT-5.6 Cyber) and was credited by Mozilla in its Firefox 159 security advisory after the team reported 81 previously unknown high-severity Firefox vulnerabilities and working exploits for 948 Linux kernel privilege escalations, 588 of them new.", "body_md": "# OpenAI’s GPT-6 Astra becomes its first model rated Critical for cybersecurity\n\nOpenAI official logo (public domain, Wikimedia Commons) — CryptoBriefing brand treatment\n\nA companion model, GPT-6 Cyber, is already working with security teams at Mozilla, Palo Alto Networks, and Cisco\n\n[OpenAI](https://cryptobriefing.com/markets/openai/) says its new GPT-6 Astra model has crossed the “Critical” cybersecurity capability threshold under the company’s Preparedness Framework. It is the first model in OpenAI’s lineup to earn that rating.\n\nAlongside Astra, the company has a specialized sibling called GPT-6 Cyber. OpenAI says it is serving security teams at Mozilla, Palo Alto Networks, and [Cisco](https://cryptobriefing.com/markets/cisco/).\n\n## What Astra can actually do\n\nAccording to the research findings, GPT-6 Astra can autonomously find and exploit zero-day vulnerabilities across complex systems. The model can also build full cyberattack plans starting from high-level goals.\n\nOn the ExploitBench evaluation, Astra posted a score of 100%. Its predecessor, GPT-5.6 Sol, scored 78.5% on the same test.\n\nOpenAI’s internal testing ran from September 1-4, 2026. During that window, Astra turned up two zero-day vulnerabilities that were previously unknown.\n\nThose two flaws will be disclosed to the maintainers of the affected software, according to the findings.\n\n## GPT-6 Cyber goes to work\n\nOpenAI released GPT-6 Cyber on November 5, 2026. It is a specialized variant built for controlled vulnerability research and exploit validation.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nAccess is limited to vetted customers only.\n\nOn the ExploitGym benchmark, GPT-6 Cyber scored 48.5%. The previous version, GPT-5.6 Cyber, scored 32.1%.\n\nThe research team reported finding 81 previously unknown high-severity vulnerabilities in Firefox.\n\nThe team also reported developing working exploits for 948 Linux kernel privilege escalations, with 588 of those being new.\n\nMozilla credited OpenAI in its Firefox 159 security advisory for the vulnerabilities found.\n\nLee Klarich, who serves as CPO and CTO of Palo Alto Networks, praised the model’s effectiveness for defensive work. Mozilla engineer Brian Grinstead did the same.\n\n## Why the Critical rating matters\n\nOpenAI’s Preparedness Framework is the company’s internal system for grading how dangerous a model’s capabilities could be in sensitive areas. Cybersecurity is one of those areas.\n\nReaching “Critical” means OpenAI itself judges Astra’s abilities to be at the top of its risk scale.\n\nOpenAI’s answer, at least for the Cyber variant, appears to be gatekeeping. Restricting access to vetted customers and routing discoveries through responsible disclosure is how the company is trying to keep the defensive upside while limiting the offensive downside.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity", "canonical_source": "https://cryptobriefing.com/openai-gpt-6-astra-critical-cybersecurity/", "published_at": "2026-11-12 18:48:26+00:00", "updated_at": "2026-11-12 18:49:49.559380+00:00", "lang": "en", "topics": ["ai-safety", "artificial-intelligence", "large-language-models", "ai-policy"], "entities": ["OpenAI", "GPT-6 Astra", "GPT-6 Cyber", "Mozilla", "Palo Alto Networks", "Cisco", "Lee Klarich", "Brian Grinstead"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity", "markdown": "https://wpnews.pro/news/openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity.md", "text": "https://wpnews.pro/news/openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity.txt", "jsonld": "https://wpnews.pro/news/openais-gpt-6-astra-becomes-its-first-model-rated-critical-for-cybersecurity.jsonld"}}