History
GPT-6 Cyber is announced for security work
OpenAI announced GPT-6 Cyber on 5 November 2026 as a specialised model for software security.

In software security, you want to follow a reported problem through to its repair. On 5 November 2026, OpenAI announced GPT-6 Cyber. The announcement fits the emergence of separate models for security work. The Wikipedia source used records the announcement under that date.

The historical point is specialisation within a model family. Here, the name Cyber refers to a separately announced model. This source entry provides no basis for saying every business could use it at the time. Nor does the announcement prove that an existing application became secure as a result.

For a small business, it helps to make your own question concrete. Do you want to understand a maintenance report better, or does a developer need to fix a reported bug? Choose the desired outcome first and assign someone to assess it. Keep sensitive details outside an initial exercise.
For repair work, request a check showing that the reported behaviour has changed. Then assess whether the application still does what the user expects. This keeps a model a tool within a clear task. Want to define that task well? Discuss your software question with Denkbaar.