Oct 30, 2026
OpenAI introduces enterprise privacy protections across ChatGPT, Codex and the API, including a Private Inference preview using confidential computing and verifiable controls for model processing.
Private Inference is a preview for security teams to assess, with an interest-registration route rather than general availability. Zero Data Retention and Private Safety Processing have workflow-specific eligibility and controls. The announcement combines new preview capabilities with existing enterprise protections.