Zero Data Retention

Zero Data Retention is a policy where an AI service deletes the questions users send and the answers it gives right after processing, rather than keeping them on its servers.

Think of it like a post office that delivers letters but keeps no copies. Companies in fields like law, healthcare, and finance—where leaked conversations can cause serious problems—have long required this kind of guarantee when using AI providers' APIs, and providers like OpenAI have agreed to it, promising not to retain the data.

The catch is that if nothing is kept, there's also no way later to check whether the AI was used for something harmful. That's why recent efforts try to solve this dilemma by adding a separate screening mechanism that automatically flags risky signals without any human ever looking directly at the conversation content.

The article reports that "OpenAI said it will continue its zero data retention policy for eligible API customers even for its frontier models." What's easy to misread here is that this isn't a new feature—it's a reaffirmation of an existing policy. What is new is a separate safety-check mechanism designed to detect risk without storing any data.