Procurement should seek explicit limits on supplier data use and a clear description of the AI functionality it is buying. The agreement should also address material changes to those features and establish how the supplier will charge for them. Work with legal and security to define the protections your organization needs for its intended use case.

At ProcureCon IT Sourcing 2026, Jonathan Knowles, Senior Category Manager at FICO, encouraged buyers to negotiate commercial terms before teams became accustomed to an AI tool. He advised procurement to "at least get some baseline pricing set in place" before a later purchase or renewal. Once engineers have adopted a feature and want to keep it, procurement can face greater internal pressure to accept the supplier's pricing and conditions.

In another 2026 discussion, a procurement practitioner recommended defining ownership of AI-generated content and negotiating how suppliers would return customer data after termination. She emphasized that buyers should receive their data in a readable format they could use after leaving the service.

Before signing, document the approved data uses and require the supplier to describe the included AI features. Review its pricing model with IT finance, including what happens after a trial ends, and agree on how the supplier must handle material changes. Ask legal to record the negotiated protections in the agreement, with clear responsibilities for reviewing changes and resolving disputes.

Join ProcureCon IT Sourcing to compare how procurement leaders negotiate AI contract protections and manage supplier commitments.