Offering Zero Data Retention for frontier models
OpenAI is retaining Zero Data Retention for eligible API customers and previewing private safety processing, relevant to teams sending sensitive context to frontier models.
OpenAI reaffirmed **Zero Data Retention** for eligible API customers and previewed **Private Safety Processing** for safety checks designed to preserve data privacy.
Builders sending proprietary code or customer data to frontier models should verify eligibility and assess whether the proposed processing model meets their retention requirements.
OpenAI reaffirmed **Zero Data Retention** for eligible API customers and previewed **Private Safety Processing** for safety checks designed to preserve data privacy. Builders sending proprietary code or customer data to frontier models should verify eligibility and assess whether the proposed processing model meets their retention requirements. The material provides no eligibility criteria, technical architecture, rollout date, or supported models for Private Safety Processing.
This adds a provider-level retention option for sensitive API workloads and previews privacy-preserving safety checks, narrowing one barrier to sending proprietary code or customer data to frontier models. It does not establish deployability yet: eligibility, architecture, model coverage, and timing remain unspecified, and retention controls alone do not satisfy broader governance requirements.