OpenAIfeatureOpenAI Private Intelligence
OpenAI's set of data controls for businesses that want to use its top models on sensitive work, including a new Private Inference preview in the API.
Details
- Groups OpenAI's business protections under one name: access restrictions, data ownership, agent action controls, and data security.
- Private Inference, in API preview, uses confidential computing to protect prompts and responses while the model is processing them.
- Zero Data Retention is available for eligible API workflows, meaning prompts and responses are not kept after processing.
- Admins set what data agents can reach, which tools and actions they can use, and when a person has to approve.
- Businesses own their inputs and outputs. OpenAI points to SOC 2 Type 2, ISO certifications, HIPAA support through BAAs, and FedRAMP for government offerings.
Best for
Getting security and legal sign-off to use OpenAI models on confidential data.
When to use it
A rollout is stuck on questions about data retention, ownership, or what agents are allowed to touch.
Who benefits
IT, security, and procurement teams evaluating ChatGPT Enterprise or the OpenAI API.
Join the newsletter to get ahead with AI.
I share strategies, new modules, free guides, and tools to get you better at using AI in your work.
Leading a team? See team programs →