What Is Agentic AI in Procurement? Gartner Explains

A Gartner analyst explains what agentic AI in procurement really means, and why it matters. At Zycus Horizon US Edition, Alex Brady, Director Analyst at Gartner, cuts through the hype around AI in procurement and gives procurement leaders a common language for it. He separates the three techniques that matter, machine learning predicts, generative AI creates, and agentic AI acts, and draws the line most teams miss: an AI agent is not the same as agentic AI. Using Gartner data, he covers where AI investment is actually paying off, the 21 to 27 percent returns organizations saw in 2025, why the biggest spenders are not always the biggest winners, and Gartner predictions that by 2030 AI will orchestrate procurement in up to 30 percent of organizations and assist 90 percent of sourcing events. He closes on what makes it work: data, governance, AI literacy, and designing procurement to work with AI, not just use it. Chapters: 0:00 Gartner’s 2030 prediction for AI in procurement 0:43 Out of the POC era, into augmentation and automation 1:50 The three AIs: machine learning, generative AI, agentic AI 3:39 The AI maturity curve 5:32 What 80% of executives say about AI returns 6:23 How much are they really spending 8:42 The ROI: 21 to 27 percent in 2025 9:54 What agentic AI actually means 11:32 AI agent vs agentic AI 14:13 Agentic AI in action 16:23 From AI assistants to a full agentic ecosystem 17:28 Capability vs reliability 18:20 Why it starts with data and governance 19:50 AI literacy as the success factor 20:44 Category management: six weeks to three hours 22:35 Sourcing: 90% AI-assisted by 2030 23:26 Where to start with agentic sourcing Learn more: Zycus PLAN 2026 (Global Virtual Event, 18 November): https://tinyurl.com/hvsjnrhh