AI-Enabled Strategic Sourcing Complexity Reduction in Public Sector

Thesis Abstract

The public sector routinely confronts procurement complexity arising from fragmented supplier ecosystems, opaque governance structures, and evolving regulatory demands, all of which impede timely and value-driven sourcing decisions. This study investigates how artificial intelligence (AI)-driven solutions can reduce strategic sourcing complexity within public procurement, with the aim of enhancing transparency, efficiency, and value realization. The research objectives are to (1) identify AI-enabled mechanisms that disrupt conventional sourcing complexity, (2) evaluate the impact of these mechanisms on procurement cycle time, cost savings, and supplier diversity, (3) examine governance and risk considerations associated with AI adoption in public sourcing, and (4) develop a contextual framework for deploying AI-infused strategic sourcing in public sector organizations. A mixed-methods research design integrates quantitative and qualitative strands to capture measurable performance improvements and contextual insights. The population comprises public sector procurement units across five metropolitan municipalities and three regional government departments. A purposive sample of 40 procurement offices is selected, yielding 200 completed procurement process records and 60 semi-structured interviews with senior procurement officers, category managers, and IT governance leads. Data collection employs (i) structured procurement process dashboards and archival records to quantify cycle time, bid rigour, cost savings, and supplier concentration, and (ii) semi-structured interviews and focus groups to elicit experiences, perceived complexity drivers, and governance controls. Instrument validity is established through expert review by seven procurement scholars and four AI ethics specialists, with reliability assessed via Cronbach’s alpha for survey items (target >0.80). Analytical methods consist of (i) multivariate regression to identify the relationship between AI-enabled features (predictive analytics, automated supplier segmentation, dynamic risk scoring, and conversational AI-supported due diligence) and performance outcomes (cycle time, total cost of ownership, and supplier diversity), controlling for organization size and sectoral constraints; (ii) interrupted time-series analysis to evaluate changes in performance trajectories before and after AI deployment; (iii) thematic analysis of interview transcripts to extract recurrent themes related to adoption barriers, governance, and perceived risks; and (iv) structural equation modeling to test the theoretical pathways linking AI capabilities, process complexity reduction, and procurement performance. The study integrates information processing theory and the resource-based view to interpret how AI-enabled information processing reduces cognitive and organizational complexity and how this translates into competitive public value. Expected findings indicate that AI-enabled features significantly shorten procurement cycles by 18–25%, improve bid quality through enhanced supplier pre-qualification, and increase supplier diversity by 10–15% in selected categories. Dynamic risk scoring is anticipated to reduce non-compliant supplier engagement and lower incident-driven delays. However, findings may reveal that governance and ethical considerations—such as algorithmic bias, data quality, and transparency requirements—moderate the strength of these effects, necessitating robust governance artifacts, explainable AI interfaces, and continuous auditing. The study contributes to knowledge by (i) operationalizing a comprehensive AI-enabled strategic sourcing framework tailored to public sector contexts, (ii) providing empirical estimates of AI’s impact on key procurement performance metrics, and (iii) detailing governance configurations that balance efficiency with accountability and public trust. The theoretical contribution lies in integrating information processing theory with the resource-based view to articulate a mechanism by which AI reduces complexity in strategic sourcing and generates organizational capabilities in public sector procurement. Practically, the research informs policy-makers and procurement leaders on selecting AI-enabled modules, designing governance controls, and sequencing implementation to maximize value while mitigating risks. Recommendations include establishing standardized data governance protocols, adopting explainable AI dashboards for decision transparency, creating cross-agency AI steering committees, and piloting AI capabilities within high-volume, low-variance categories before scaling. The study concludes that when deployed with rigorous governance, transparent accountability mechanisms, and continuous performance monitoring, AI-enabled strategic sourcing can meaningfully reduce structural complexity in the public sector and deliver measurable improvements in efficiency, integrity, and value realization.