GPT-6 Cyber and Gemini Carbon reach vetted security teams first
OpenAI’s GPT-6 Cyber joins Google’s Gemini 4 Carbon in a restricted rollout, with stronger reported security capabilities and access conditions that prevent treating either as an ordinary public model launch.
OpenAI introduced GPT-6 Cyber on November 5 for authorized vulnerability discovery, exploit development and security testing.
It assigns the model its Critical cybersecurity capability designation. Access requires both Daybreak Red approval and separate approval for GPT-6 Cyber; buying an enterprise contract or enabling Red does not by itself grant the model. [1][2][3]
The strongest practical evidence is that some findings have become fixes. Mozilla’s Firefox 159 security advisory lists high-severity vulnerabilities reported by OpenAI Codex Security. That independently establishes accepted findings and released remediation; it does not independently reproduce OpenAI’s broader claim that GPT-6 Cyber found 81 previously unknown high-severity Firefox vulnerabilities. [1][4]
Google announced Gemini 4 Carbon on October 28, beginning with selected cyber defenders and extending access through Fairwind. Its announcement described wider developer and consumer availability as forthcoming. A founder choosing a security research stack therefore faces two decisions: whether the organization qualifies for access, and whether the resulting workflow produces validated fixes within a defined scope. [5][6]
Sources & details 6 sources
Discuss in your voice app
Take the reporting, visual explanations and original sources into a conversation. Copy the brief, paste it into your preferred assistant, then start its voice mode.
Paste the brief into your voice app to discuss this reporting.Reporting sources
- OpenAI — GPT-6 Cyber
November 5 launch and vendor-reported capabilities; benchmark and aggregate discovery claims are not independent replications.
- OpenAI — GPT-6 Cyber system card
Capability classification, access requirements and residual alignment limitations.
- OpenAI Help — Daybreak access
Separate model approval, organization eligibility, internal-use restrictions and security setup.
- Mozilla — Firefox 159 security fixes
Independent affected-software maintainer credits OpenAI findings and documents fixes; not a replication of the model's overall benchmark or discovery count.
- Google — Gemini 4 Carbon
October 28 announcement specifies staged access before broader product release.
- Google DeepMind — Fairwind
Program context for vetted defensive access; not proof of unrestricted model availability.
Images & figures
AI-generated editorial illustration of powerful security research restricted to an authorized scope; not a product architecture or evidence of infallible boundary adherence.
Euangelion · AI-generated editorial illustration · Source · Usage terms