Circular procurement is becoming harder to manage as firms face volatile material markets, tighter sustainability requirements and more fragmented supply networks. A study published in the Business Strategy and the Environment journal argues that one of the most promising responses is not conventional analytics, but agentic AI: systems that can scan markets, assess options and execute procurement actions with limited human intervention.
Drawing on resource orchestration theory,...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
the authors examine whether adopting agentic AI in industrial purchasing improves circular procurement performance, and how that happens. Their central finding is that firms gain most when the technology is supported by three organisational conditions: disciplined resource scanning and evaluation, tightly integrated cross-functional procurement systems, and established routines for collaboration with suppliers. In other words, the value of agentic AI depends less on the technology alone than on the surrounding capability structure.
The study is framed against a broader shift in procurement practice. Circular procurement requires buyers to source recyclable and remanufactured materials, manage reverse flows, and align procurement with engineering, logistics and sustainability teams. Traditional digital tools have improved visibility and traceability, but they remain largely human-directed. Agentic AI, by contrast, is presented as a more autonomous layer of decision support and execution, capable of identifying risks, simulating trade-offs and triggering workflow changes within governance limits.
The authors surveyed managers in India, a context they describe as particularly relevant because of rapid digitalisation, expanding manufacturing capacity and rising attention to circular economy policies. Responses came from firms across sectors including IT services, automotive, chemicals, logistics, pharmaceuticals and renewables. The analysis used covariance-based structural equation modelling and the PROCESS method to test a serial mediation model linking agentic AI adoption to circular procurement performance through resource acquisition, resource integration and collaborative innovation.
The results support both a direct and an indirect relationship. Agentic AI adoption was associated with better circular procurement performance on its own, but the stronger effect ran through a step-by-step capability chain. First, firms used the technology to improve resource acquisition, such as identifying circular suppliers, scanning external markets and securing relevant data and technologies. Next, these resources were embedded into procurement routines and cross-functional systems. Finally, the integrated capability base supported collaborative innovation with suppliers, including joint experimentation and co-development of circular solutions.
All of the proposed mediation paths were significant, including the full sequence from agentic AI to resource acquisition, resource integration, collaborative innovation and then circular procurement performance. The control variables, firm size and firm age, were not significant, suggesting that the pattern held across organisations of different scales and maturities.
The paper’s theoretical contribution is to extend resource orchestration theory into the realm of autonomous AI. It argues that agentic systems do not merely support managers; they also partially take on orchestration tasks themselves, creating a hybrid model in which human judgement and machine autonomy work together. Practically, the authors say procurement leaders should not treat agentic AI as a standalone tool. They should first build the processes and governance needed to source circular inputs, then connect the technology to enterprise systems, and only then use it to deepen supplier collaboration and innovation.
The study also notes limitations. It is cross-sectional, relies on single-informant perceptions and focuses on one developing economy. Future research could compare countries, use objective performance measures and distinguish between different levels of AI autonomy. Even so, the paper’s message is clear: in circular procurement, agentic AI is most powerful not as a replacement for orchestration, but as an engine that expands it.
Source: Noah Wire Services
