OpenAI is highlighting an important privacy win for organizations adopting advanced AI: Zero Data Retention for eligible API customers using frontier models. This means qualifying customers can benefit from powerful AI capabilities without having their inputs or outputs stored by OpenAI.
For businesses, developers, and institutions working with sensitive information, this can help reduce barriers to responsible AI adoption. Stronger privacy controls are especially valuable in sectors where data governance, confidentiality, and compliance are central to deploying new technology.
Privacy and safety moving together
OpenAI also previewed Private Safety Processing, an approach designed to support advanced AI safety systems without compromising customer data privacy. That direction matters because frontier AI needs robust safeguards, but organizations also need confidence that their data remains protected.
- Greater trust: Privacy-forward infrastructure can help more organizations explore AI responsibly.
- Enterprise readiness: Zero Data Retention supports use cases with higher security and compliance needs.
- Balanced progress: Private Safety Processing points to a future where safety and privacy improve together.
This is a practical step toward making frontier AI more usable in real-world environments where privacy is not optional. By pairing powerful models with stronger data protections, OpenAI is helping broaden safe and trusted access to advanced AI.