OpenAI is preparing to preview and then unveil GPT-6 Cyber, a new cybersecurity-focused model, within weeks. Multiple outlets report that the model is being trialed by a limited group of users through OpenAI’s advanced access program. The launch is also discussed as aligning with OpenAI’s broader developer showcase plans around its annual DevDay event on September 29.
Reporting from different sources describes a two-part release. In addition to GPT-6 Cyber, OpenAI plans a new product intended to help customers deploy the model more securely and with greater automation. One outlet says the deployment tool’s name has not been announced, but that it is designed to streamline security workflows such as handling vulnerabilities and patching.
While the specifics of features and timing are not fully consistent across accounts, the overall themes match: a cybersecurity model aimed at security operations, paired with an accompanying product to support safer, easier deployment. Sources characterize this as an expansion of OpenAI’s existing approach of pairing core models with an enabling product layer.
