By independently carrying out intricate activities like partner sourcing, negotiation, risk management, and dynamic contract lifecycle oversight, agentic artificial intelligence (AAI) is revolutionizing procurement and offering previously unheard-of efficiency, cost savings, and strategic value. However, technology presents new hazards outside traditional frameworks due to its independent, adaptable, and occasionally opaque nature. These include operational risks (integration failures, unintended behaviors), cybersecurity risks (prompt injections, AI worms), ethical/societal risks (algorithmic bias, accountability gaps, job displacement), data risks (integrity, privacy), and regulatory/compliance risks in changing legal environments. With a focus on proactive identification (e.g., attacker testing), mitigation through technical safeguards (e.g., explainable AI, robust data governance), organizational controls (e.g., transparent governance, incremental rollout), and human-centric approaches (e.g., human-in-the-loop mechanisms), this chapter offers a structured, multi-layered risk management framework. Adaptive governance, transparent accountability, and ongoing monitoring are essential for maximizing the advantages of AAI while guaranteeing its moral, robust, and reliable incorporation into procurement ecosystems.