A consortium of independently governed public libraries wants to deploy a shared AI-powered catalog assistant that searches across every member library's collection. Each library maintains its own board and its own policies about its catalog data. What question must the consortium resolve before deploying the shared assistant?

Because each library in the consortium is independently governed with its own board and its own catalog policies, deploying a shared assistant means resolving how data ownership and usage rights work across organizations that don't report to a common authority. That agreement needs to cover who controls how a member's catalog data is used, what happens if a library later withdraws, and how disputes over data use get resolved, since none of that is implied just by wanting to build the tool together. Deciding which library hosts the user interface is a technical hosting detail that matters far less than whether the underlying data-sharing relationship is even settled. Allocating staff hours for patron support is an operational staffing question that assumes the assistant is already cleared to launch. Choosing a typeface and color scheme is a cosmetic decision with no bearing on whether the consortium can legally and operationally combine each library's catalog data. This applies specifically to shared systems spanning independently governed organizations; a single library building its own internal catalog tool wouldn't face this governance question. A concrete check is confirming a signed interlibrary agreement names each member's data rights and exit terms before the shared assistant goes live.