Oracle OCI Generative AI
Oracle says OCI Generative AI does not retain customer inference inputs or outputs, does not share them with third-party model providers, and does not use fine-tuning data to improve general OCI Generative AI use cases.
Customer-directed fine-tuning uses training data stored in the customer tenancy to build that customer's custom model. Oracle says it does not retain that data beyond the job or use it for general improvement.
No opt-out is required for the stated inference and general-training posture. Customers choose whether to run fine-tuning and manage training data in their tenancy.
Prompts, responses, fine-tuning data, and custom models are not shared with named third-party model providers according to the data-handling page.
Base inference inputs and outputs are not stored by OCI Generative AI. Fine-tuning data is deleted from the fine-tuning cluster when the job completes. Separate Generative AI Projects can retain responses and conversations for a customer-configured period up to 30 days.
The reviewed data-handling documentation does not state a general human-review process for inference content.
Foundational reference pages for teams adopting, deploying, and operating production AI with inspectable controls.