Stan Garber is the President and Co-Founder of Levelpath, a premier AI procurement platform.getty​Enterprise CTOs and CIOs have a new assignment: Agentify everything. Customer service! Lead scoring! Employee onboarding! Even procurement, a process they’ve typically treated as “not my job,” is on that list. As tech leaders take more ownership of procurement, they’ll learn that AI agents and a supplier graph are directly tied to IT resilience, risk visibility and speed across the enterprise. ​Many organizations think of vendor risk as strictly a procurement issue, something to manage after the contract is signed. That’s not the whole picture. Supplier data (which I've found is risk’s top informant) falls under the purview of tech leadership, even if CTOs and CIOs haven’t seen it that way. Tech spend is often one of the largest variables in indirect procurement that I've seen in enterprises. CTOs and CIOs have to clear every tech vendor before onboarding. Supplier data is still data, after all. ​The CTOs and CIOs who’ve started deploying AI agents to help with procurement processes have seen their impact on tech-related risk. I know a CIO at a large construction company whose supplier didn’t deliver on a contractually committed SSO rollout. The CIO pointed an AI agent at the contract to understand their exposure. The agent surfaced the breach, drafted an email response to the supplier and produced a formal breach letter. ​That CIO saw what other tech leaders agentifying their processes are about to discover: AI agents can help turn vendor risk visibility into an IT operational capability, one that, with the right infrastructure, can run continuously and surface issues before they become major events. ​Supplier Graphs Are IT Infrastructure ​Before tech leaders can fire AI agents at any procurement workflow, they need to build a foundation for them to work on. ​The office of the CIO owns more supplier data every year, as companies continue to inject technology into all areas of the business. The data underlying that spend, however, usually lives in ERPs, contract repositories and email threads nobody can really search through. What CIOs and CTOs need is a supplier graph shared with finance, security and legal, where every sourcing request and contract flows into one place. ​They can start to create this supplier graph by making the procurement front door accessible enough for the whole company to use. When everyone routes through the same system, all the information CIOs and CTOs might need to make a pivot, like supplier information, approval chains and renewal deadlines, finally becomes searchable. ​AI Agents Can Spot Vendor Risks ERPs Can’t See​When a disruption like a data breach or supply chain failure hits the business, data tends to be the first problem. How many CTOs and CIOs can easily answer these questions: Who do you actually buy from? How exposed are you? What vendor can you call instead? When you don’t know the answers to these questions, you can’t act quickly. ​Agents running on a supplier graph can surface information obscured by fragmented records. They can find, for example, their suppliers’ suppliers, three “different” vendors that roll up to one parent or a cluster of critical suppliers in a country whose pricing will soon be affected by new tariffs. This level of entity resolution and sub-tier supplier mapping is beyond what most ERPs can do. ​Illuminating the concentration risk in your direct spend and classifying vendors so users can find them easily is a great win for leaders. Who among us hasn’t stared at a supplier master and wondered what half of those vendors do?​You Can Start Small With AI Agents For Procurement​Your first AI agent doesn’t need to be sophisticated; a contract review agent that flags non-standard terms is enough. ​The actual work involves getting the supplier graph structured so agents know what information to reference, figuring out what decisions should be automated and which require human judgment, setting guardrails and building a feedback loop so agents improve over time. ​The goal is to build agents that handle whole steps of the procurement process independently. Reviewing contracts against standard playbooks, scoring supplier responses, routing approvals and flagging upcoming renewal deadlines are all reasonable candidates. ​Once you’ve got them up and running, you can measure whether your pilot is working by picking one or two high-spend categories or regions and watching the percentage of spend on contract and speed. Both track procurement resilience. An industrial equipment company I work with found a $3.5 million pricing discrepancy on a single contract in week one of deploying AI agents for procurement. Another cut its contract review cycle times by half. ​This Isn’t Just A Tool For The Procurement Department ​Procurement is the one major spend domain without a legitimate system of record. Closing that gap takes an investment in AI infrastructure, and that decision belongs to CIOs and CTOs. You’ve got the data. Now you need to connect it in a way agents can use to make your tech spend more resilient and risk-proof. ​​​Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?