OpenAI launches GPT-6 Cyber, its first model rated Critical for cybersecurity
OpenAI has released GPT-6 Cyber on 5 November, restricting it to approved Daybreak Red customers via the API and Codex. The company rates it Critical for cybersecurity under its Preparedness Framework, up from High for GPT-5.6 Cyber, and cites results that remain unverified independently.
OpenAI released GPT-6 Cyber on 5 November 2026, a security-focused model available initially only to selected, approved customers of its Daybreak Red programme. Access is through the API and Codex, and OpenAI has said it plans to widen availability to all Daybreak Red customers soon.
The company classifies the model as Critical for cybersecurity under its Preparedness Framework. That is a step up from the High rating given to its predecessor, GPT-5.6 Cyber, and signals that OpenAI regards the system as capable of materially assisting offensive security work.
OpenAI reports a 48.5 per cent score on ExploitGym, 948 Linux exploits and 81 new Firefox vulnerabilities found. These are company claims and have not been independently verified. No public pricing or general-availability date has been given.
Daybreak Red sits alongside Daybreak Blue, the defensive tier, and a $1 billion grant programme intended to subsidise access for frontline defenders, according to coverage of the launch.
Why it matters
Gated release is becoming the template for models with strong offensive capability: the vendor vets users instead of publishing the model openly. For security teams, that means access depends on verification, not just a credit card.
The Critical rating also gives regulators and boards a concrete reference point when judging how far frontier systems have moved into cyber operations.
Indicops view
Indian CERTs, banks and critical-infrastructure operators should ask now how they would qualify for vetted-access programmes, and should treat vendor-reported exploit counts as claims to test in their own environments. Defensive tooling plans should assume attackers will reach similar capability within months.
More in Models
Back to the newsroomGoogle launches Gemini 4 Hydrogen, a $1.50-per-million-token model for coding and agents
Google has released Gemini 4 Hydrogen, a faster and cheaper tier of its Gemini 4 family aimed at everyday coding and agentic work. It costs $1.50 per million input tokens and $7.50 per million output tokens across a 1M-token window, and is rolling out through the Gemini API, AI Studio and consumer products.
Alibaba opens a two-week preview of Qwen4-Max with a 1M-token context, but no benchmarks or weights
Alibaba Cloud has opened a time-limited preview of Qwen4-Max, its next flagship model, through the Token Plan only. It offers a 1M-token context and up to 393,216 output tokens, but the documentation lists no pricing, benchmarks or downloadable weights, and the model retires on 25 November.
Anthropic releases Claude Fable 5.2 and Mythos 5.2, 30 per cent faster and 20 per cent cheaper
Anthropic has released Claude Fable 5.2 for general use and Mythos 5.2 for vetted cyber and life-sciences organisations. The company says output is up to 30 per cent faster than the 5.1 generation, at $8 per million input tokens and $40 per million output tokens, with lower rates of safeguard blocks.