Safety overview: GPT-6 Astra
GPT-6 Astra is OpenAI's first model rated Critical for cybersecurity capability under its Preparedness Framework, a material consideration for security-sensitive agent access and controls.
OpenAI calls **GPT-6 Astra** its most capable broadly deployed model. It is also the company's **first model** to reach the **Critical cybersecurity level** under its Preparedness Framework.
Builders giving agents access to code, credentials, networks, or security tools should treat model capability as part of the threat model and revisit permissions, isolation, and audit controls.
OpenAI calls **GPT-6 Astra** its most capable broadly deployed model. It is also the company's **first model** to reach the **Critical cybersecurity level** under its Preparedness Framework. Builders giving agents access to code, credentials, networks, or security tools should treat model capability as part of the threat model and revisit permissions, isolation, and audit controls. The supplied overview does not explain the evaluation, mitigations, deployment restrictions, or practical meaning of the Critical rating. It supports caution, not a quantified estimate of application risk.
The Critical cybersecurity classification raises the security stakes of selecting Astra even though it does not quantify application risk. It strengthens the case for treating the model as untrusted and enforcing authority outside it; without evaluation and mitigation details, builders should revisit permissions, isolation, credential handling, and auditing rather than infer either safety or danger precisely.