Executive Summary
AI Gateway has launched new compliance features, centered around team-wide Zero Data Retention (ZDR), to simplify and enforce data privacy policies for organizations using multiple AI models. The update allows Pro and Enterprise teams to activate a "no-code" ZDR policy across all requests from a central dashboard, automatically routing traffic only to compliant providers. This shifts the burden of managing fragmented data policies from individual developers to a centralized infrastructure, aiming to eliminate manual, error-prone configuration.
Key Takeaways
* Team-Wide Zero Data Retention (ZDR): A new dashboard setting for Pro and Enterprise teams that enforces ZDR on all AI requests without requiring any code changes.
* Per-Request Controls: Developers can still enforce ZDR or disallow prompt training on a case-by-case basis for specific sensitive workflows via API options.
* Automatic Compliant Routing: The gateway automatically routes requests only to model providers with which a ZDR agreement is in place, including select models from OpenAI, Anthropic, and Google.
* Disallow Prompt Training: A separate, more granular option prevents providers from using prompt data for model training, which is automatically included under the broader ZDR policy.
* Audit Trail: Each API response includes metadata detailing how routing decisions were made and which policies were enforced.
Strategic Importance
This update positions AI Gateway as a crucial infrastructure layer for enterprises, addressing the significant business challenge of data privacy and compliance in multi-model AI development. By abstracting this complexity, the platform becomes more attractive to security-conscious customers and increases its strategic value beyond simple model routing.