OpenAI keeps Zero Data Retention for frontier models while previewing Private Safety Processing to detect risks across AI interactions without exposing user content

OpenAI says it will keep offering Zero Data Retention for its frontier models. The company explains that as AI handles longer, more autonomous work and delivers greater business value, safety systems need to identify risks across related interactions. To address this, OpenAI is previewing Private Safety Processing, designed to improve safety without giving OpenAI personnel access to the underlying content.

✨ AI summary of the linked source — not editorial content.

Explore the topic

• Risks
• OpenAI
• AI