OpenAI Developers said in a post on X that records stay in the customer's environment, where an automated safety agent reviews them. The post said OpenAI personnel are not able to access the content.
OpenAI Developers say they have launched Private Safety Processing for API requests
Feature keeps records in the customer's environment, company says

OpenAI Developers said in a post on X that OpenAI has launched Private Safety Processing, a feature that lets it review related API requests for safety without retaining the prompts or responses involved.
The Daily Edition
The front page in your inbox each weekday morning: the lead, the picks, and why each one earned its place.
§
Analysis
Why This Matters
- Developers using OpenAI's API may be able to have safety checks performed without the prompts and responses being stored by OpenAI, a potential answer to privacy concerns about AI-powered services.
- The announcement is notable for enterprises handling sensitive data, though the company's claims are not independently verified.
- It reflects a continuing effort by AI providers to separate safety review from data retention.
Background
AI companies offering APIs typically run safety filters on inputs and outputs, and such review has often involved storing or accessing user content. This feature is described as an alternative that keeps records in the customer's own environment while still allowing an automated review.
What to Watch
- Whether OpenAI publishes more detail on how the automated safety agent operates.
- Whether independent audits or testing confirm that prompts and responses are not retained.
- How widely the feature is adopted and whether it is offered across more API products.
Sources
- OpenAI launches Private Safety Processing for API requests · OpenAI Developers on X
Checked
7 claims in this story checked against 1 source: 6 supported.
Show each claim
- Partly supportedOpenAI Developers say they have launched Private Safety Processing for API requests.OpenAI Developers on X: “OpenAI launches Private Safety Processing for API requests.”
- SupportedOpenAI has launched Private Safety Processing for API requests.OpenAI Developers on X: “OpenAI launches Private Safety Processing for API requests.”
- SupportedThe feature keeps records in the customer’s environment, the company says.OpenAI Developers on X: “Records stay in your environment”
- SupportedOpenAI Developers said in a post on X that the feature lets OpenAI review related API requests for safety without retaining the prompts or responses involved.OpenAI Developers on X: “Private Safety Processing lets OpenAI review related API requests for safety without retaining your prompts or responses.”
- SupportedRecords stay in the customer’s environment.OpenAI Developers on X: “Records stay in your environment”
- SupportedAn automated safety agent reviews the records.OpenAI Developers on X: “where an automated safety agent reviews them”
- SupportedOpenAI personnel are not able to access the content.OpenAI Developers on X: “without OpenAI personnel being able to access your content.”