OpenAI Reaffirms Zero Data Retention and Previews Private Safety Processing
OpenAI has reaffirmed its Zero Data Retention (ZDR) policy for eligible API customers and introduced Private Safety Processing. This new feature allows the company to perform advanced AI safety checks on prompts and responses without storing or compromising user data. This update addresses critical data privacy and security concerns for enterprises adopting large language models (LLMs). By ensuring that sensitive prompts and outputs are not retained during safety evaluations, OpenAI lowers the barriers for highly regulated industries to integrate frontier AI models. Under the Zero Data Retention policy, OpenAI does not retain prompts or model responses after a request is processed. The previewed Private Safety Processing mechanism ensures that safety checks are executed transiently, preventing storage risk at the provider layer.
## BACKGROUND
Enterprise adoption of AI models is often hindered by strict data compliance laws, such as GDPR, which require organizations to control where and how their data is stored. While OpenAI offers Data Processing Addendums (DPAs) to support compliance, standard API usage may still involve temporary data logging for safety monitoring. Zero Data Retention policies eliminate this logging, but historically made it harder to perform real-time safety checks, a gap that Private Safety Processing aims to bridge.