“OpenAI has introduced GPT-6 Cyber, a new AI model designed specifically for cybersecurity research, including finding vulnerabilities, developing working exploits, and helping fix security flaws." Secondary claims in the same post: (a) "the model identified 81 previously unknown high-severity vulnerabilities in Firefox" (b) "developed…”

The model exists and the announcement is real. OpenAI published a dedicated announcement page on its own index describing GPT-6 Cyber as "a major leap in security research: more capable at finding and fixing vulnerabilities, and far more reliable staying within the boundaries researchers set". Its official X account announced it the same day, stating that the model powered the Linux kernel research announced the previous day and that security teams at companies including Mozilla, Palo Alto Networks, and Cisco are already using it. On purpose and scope, OpenAI's own framing matches the claim. The announcement describes it as OpenAI's cybersecurity model for finding exploitable vulnerabilities and developing fixes, trained to stay within authorized scope. Reporting describes it as a version of GPT-6 Astra trained for security work, supporting bug discovery, exploit development and authorized security testing with new safeguards. On the Critical classification, the claim is confirmed by two primary vendor artifacts. The announcement states: "Based on these results, we classify the model as Critical in cybersecurity under our Preparedness Framework, up from High for GPT-5.6 Cyber." The system card adds the fuller picture: under the Preparedness Framework, GPT-6 Cyber receives the same capability designations as GPT-6 Astra, Critical in Cybersecurity, High in the Biological and Chemical domain, and below High in AI Self-Improvement. Context the post omits: GPT-6 Cyber is not the first OpenAI model at this level. Astra was the first OpenAI model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, announced in September 2026. On the Firefox figure, OpenAI reports that its researchers used GPT-6 Cyber to find 81 previously unknown high-severity vulnerabilities in Firefox, which Mozilla has started fixing, and that it shared the findings and proposed patches with Mozilla, which began rolling out fixes shortly after receiving the reports. One report notes the severity rating came from the affected party: Mozilla classified all 81 findings as high severity. On the Linux figure, the underlying research is a separate publication dated one day before the model launch. OpenAI states: "Our prototype identified and analyzed 2,308 candidate Linux kernel findings, dynamically reproduced 1,724 vulnerabilities, and produced 948 local privilege escalation (LPE) exploits that could allow an unprivileged local user to gain root access." The run is dated and costed: in a run in mid-October 2026, the pipeline using an upcoming cyber model dynamically reproduced 1,724 vulnerabilities and generated proofs of concept for 291 information leak vulnerabilities and 948 working local privilege escalation exploits for a commonly used Linux distribution. The spend was substantial: the run ended with 948 LPE exploits and cost $1,213,430 in API credits, or $1,279 per exploit. Human review was part of the process: OpenAI said its researchers reviewed all of the generated exploits and patches before reaching out to Linux kernel maintainers. On access, the gating claim is confirmed. GPT-6 Cyber will first be available to a limited number of customers in OpenAI's Daybreak Red program, which uses more advanced or nonpublic cybersecurity models for offensive red teaming, and it is initially available through the API and Codex to select Daybreak Red customers. The Daybreak Red tier predates this model: Daybreak Red provides access to OpenAI's purpose-trained cybersecurity models for authorized vulnerability research, exploit validation, and security testing, a tier introduced alongside GPT-5.6-Cyber in August 2026. OpenAI's own post states access is expanding over time.