Agentic Exposure Validation

Check Point moves OpenAI models into the step that decides what gets patched

Four Check Point workflows now route decisions through OpenAI's frontier cyber models, including one that builds its own exploit material, and the announcement puts a stage label on only one of them.