OpenAI Introduces Enhanced Customer Privacy Safeguards and Enterprise Zero-Data Retention

In a direct bid to capture sensitive enterprise workloads from competitors like Anthropic and Microsoft, OpenAI has unveiled a comprehensive suite of upgraded privacy and security protections for ChatGPT Enterprise and API corporate clients. The new controls provide strict, verifiable boundaries ensuring customer prompt data and model outputs are never retained or utilized for model retraining.

Central to the security update is Customer-Managed Encryption Keys (CMEK) integration, allowing corporate IT administrators to control data encryption at rest using their own cloud key management infrastructure. Additionally, OpenAI introduced confidential compute sandboxes that encrypt data in memory during inference execution.

Enterprise compliance officers can now generate real-time automated audit logs to verify data lifecycle enforcement, setting a new benchmark for corporate governance in generative AI deployments.