Skip to content
NewEraAI

AI news

OpenAI commits to zero data retention and previews private safety processing for frontier models

OpenAI has reaffirmed zero data retention for eligible API customers and previewed a new private safety processing approach for advanced AI safety. Business teams using frontier models should review which workloads are eligible and update their privacy and risk controls.

20 August 2026

Visual abstraction of neural networks in AI technology, featuring data flow and algorithms.
Photograph by Google DeepMind · Pexels

OpenAI says eligible API customers will continue to see zero data retention for frontier models. It also previews a private safety processing approach aimed at strengthening safety while keeping customer data private.

What changed

OpenAI reaffirmed zero data retention for eligible API customers. It also shared details about private safety processing for advanced AI safety that is designed not to compromise data privacy.

Why it matters for business teams

If your team uses frontier models through the API, data retention directly affects privacy risk, compliance posture, and incident response. Zero data retention can reduce how long prompts and outputs are kept, and private safety processing can help strengthen safety controls without requiring retention of customer data.

What to do next

Confirm whether your API usage is eligible for zero data retention, then update internal documentation for privacy controls and vendor risk. If you have safety monitoring workflows, map where the new private safety processing step fits into your operational review process.

Next step

Start with the free AI Opportunity Assessment.

A short, no-obligation conversation about where enquiries, hours and revenue leak today. You do not have to pick a tier to have it, and what comes out of it feeds Discover, so the first paid day starts from evidence rather than a blank sheet.