Description
AI adoption in procurement is accelerating: 80% of supply chain leaders will begin or continue gen AI rollouts within 12 months. HBR Analytic Services research shows how leaders capture value while managing the risks.
Key Takeaways
- Per Gartner, 80% of supply chain leaders and 73% of procurement leaders will begin or continue implementing generative AI within 12 months, making inaction the biggest competitive risk.
- AI value hinges on data readiness, which means, without accurate, well-governed data and safeguards for proprietary information, procurement AI tools cannot deliver reliable results.
- Workforce adoption decides AI success; since employees make the final call on AI recommendations, training, digital literacy and skills like prompt engineering are essential investments.
Executive Summary
Doing nothing on AI is now the biggest risk in procurement and supply chain management.
Adoption is accelerating fast. According to a January 2024 Gartner survey, 80% of supply chain leaders and 73% of procurement leaders will either begin or continue implementing generative AI over the next 12 months.
Yet most organizations remain in a transitional phase. Some are automating processes and forecasting demand with machine learning; others are still weighing whether to start, held back by data quality gaps, privacy concerns and workforce readiness.
This Harvard Business Review Analytic Services report, sponsored by GEP, examines how companies are putting AI, machine learning and generative AI to work across procurement and supply chain operations. It draws on Gartner survey data and insights from academics at the University of Houston and Penn State, along with practitioners at LNG producer Cheniere Energy.
The report covers where AI is delivering value today — demand forecasting, supplier risk assessment, spend analysis and process automation — and where adoption is moving more slowly. It also examines the challenges leaders must work through: getting data accurate and well-governed, protecting proprietary information from open LLMs, and preparing employees to work alongside AI tools.
For procurement and supply chain executives, the choice is no longer whether to adopt AI, but whether to lead, follow, or fall behind competitors who move first.
Frequently asked questions
Companies are wary of open LLMs learning from data that holds competitive value, such as maintenance and procurement histories. Common safeguards include partnering with AI providers offering customizable data controls, encrypting data at rest and in transit, anonymizing sensitive data before model training, and using dedicated environments that prevent data commingling.
AI follows the "garbage in, garbage out" rule, results are only as good as the underlying data. Foundational steps include implementing data integration platforms to establish a unified data strategy, investing in data cleansing and validation processes, and employing data quality management tools with strong governance.
Adaptive algorithms learn from new data and adjust forecasts in real time as sales patterns shift or disruptions occur, enabling more responsive inventory management. For supplier risk, AI evaluates performance and predicts potential issues by analyzing historical data alongside external factors such as market trends and political risks.