Generative AI (GenAI) for procurement has entered what research and consultancy firm Gartner Inc. calls the “trough of disillusionment,” a stage where excitement cools down and expectations become more realistic. While some companies that adopted the technology early are seeing benefits, others are struggling to get consistent returns, showing the need for a more careful and strategic approach.
“GenAI is proving to deliver process efficiency, better data insights, and cost savings for procurement organizations,” said Kaitlynn Sommers, senior director analyst in Supply Chain at Gartner. “However, fragmented and low-quality data across procurement systems can hinder accurate outputs, and integrating stand-alone GenAI solutions with existing platforms is often complex, due to differing technical specifications.”
Sommers added that despite these challenges, its applicability across the source-to-pay spectrum continues to drive strong interest and adoption.”
Gartner’s Hype Cycle for Procurement & Sourcing Solutions maps the stages of maturity and adoption for new technologies. Aside from GenAI, other procurement tools now in the trough of disillusionment include sustainable procurement applications, prescriptive analytics, supplier diversity solutions, and advanced contract analytics. The firm also predicts that conversational AI in procurement may become outdated before it becomes widely useful.
For those using it, GenAI in procurement is mainly applied to repetitive and time-consuming tasks such as gathering information, summarizing documents, creating workflows, and carrying out routine processes. This automation can help organizations work faster, cut costs, and free employees to focus on strategic activities like managing supplier relationships and making high-level business decisions. Some companies are also using GenAI to automatically create contracts, scope out projects, suggest suppliers, and generate “Request For” documents (RFx) using natural language commands.
However, adoption is not without hurdles. Many companies still deal with poor or incomplete data, making it hard for AI to give accurate results. Integrating AI into existing systems can be costly and complicated, and some employees are unsure about trusting AI-generated insights or worry about how it might affect their jobs. Changing long-standing processes is also difficult, and uncertainty around new regulations raises concerns about privacy, intellectual property, and trust.
Sommers warned that businesses that wait too long to explore GenAI for procurement risk falling behind competitors. Gartner expects the technology to mature and become widely productive within five years. To succeed, organizations will need to improve their data systems, choose AI solutions that fit their goals, adapt their processes, keep an eye on regulatory changes, and train staff to work effectively with AI tools.