AWS adds live governed data to Amazon Quick apps

AWS says it has introduced Live Data in Apps for Amazon Quick, allowing AI-built apps to query governed Amazon Quick Sight datasets whenever a viewer opens them instead of displaying data captured during app creation. The feature gives Amazon Quick apps current query results while keeping the controls attached to the source dataset.

After an app is published, it reruns the SQL generated during the build for each viewer. Users access apps through their existing Quick identity, interactions take place in an authenticated Quick session, and runtime access reflects the viewer’s own permissions. For apps that use live datasets, AWS says viewers must be authenticated Quick users and anonymous or public access is not supported.

AWS says Live Data in Apps applies existing row-level and column-level security when querying Quick Sight. Row-level security uses permission rules to limit which records a registered user or group can read. Column-level security controls access to specific fields and can prevent users from seeing visuals built with restricted columns. People opening the same app can therefore receive different results from the same governed dataset based on their permissions.

AWS says each viewer must give one-time consent for every dataset on first use before a live query can run. The launch post says the first app opening prompts the viewer to allow access to the datasets and that the backend rechecks consent on every query. It does not describe how a viewer can revoke that consent.

The same governance model appears in AWS’s Dataset Q&A documentation, which says natural-language dataset questions generate SQL using semantic context and that responses respect row-level and column-level security rules.