PublicInput is adding agentic AI features (chat-based AI) to products like Pulse, the AI Project Builder, and the AI Email Campaign Builder. These features are currently in beta. Trust and data protection are essential for the government organizations we serve, so here is exactly how your data is handled when you use them.

All of our AI models are hosted by Microsoft Azure in an enterprise environment. Your data is processed inside Microsoft's secure cloud. It is never sent to public or consumer AI services.

Our AI environment enforces strict data-residency requirements. Your organization's data, and any content processed by our AI features, remains within the United States.

We maintain an explicit, contractual prohibition on training models on your conversations or your data. Nothing you or your residents submit is used to train or improve any AI model.

Agentic AI features are optional and can be turned off for your entire organization at any time. An organizational administrator can manage this from Settings > General Settings (or directly from the Pulse dashboard) with a single toggle. The toggle is on by default while these features are in beta.

Turning agentic AI off only disables chat with large language models. You can still take full advantage of the data Pulse pulls together, and all other PublicInput features work exactly as before.