Integrated intelligence, autonomous operations, and trusted autonomy: AI-driven SCM performance in 2040
TL;DR
A SCOR-DS-wide, Delphi-based analysis of how AI-driven SCM performance enhancement is expected to unfold by 2040 as a layered, process-uneven configuration rather than a single linear trend shows that these trajectories operate through distinct mechanisms.
Abstract
This study aims to examine how artificial intelligence (AI)-driven advancements may enhance supply chain management (SCM) performance across SCOR-DS processes by 2040, addressing the lack of systematic, long-horizon, cross-process evidence on AI-related performance effects. A foresight-oriented real-time Delphi study assessed 13 projections developed from desk research and 12 semi-structured expert interviews, then refined in internal and expert workshops. Seventy experts rated each projection’s expected probability, impact and desirability, providing qualitative rationales. Results were analyzed using descriptive and consensus statistics, stakeholder comparisons, dissent diagnostics, qualitative coding and fuzzy c-means clustering, discussed through a complex adaptive systems lens. Experts expect substantial AI-enabled SCM performance gains by 2040. Results indicate layered AI integration. Integrated intelligence emerges as the most plausible and impactful pathway, reflecting deeply embedded, predominantly assistive AI across planning, sourcing, fulfillment and returns. The autonomous operations pathway captures more selective, contested moves toward AI-led training, shared AI infrastructures, unmanned production and autonomous mass customization. Trusted autonomy is less directly impactful yet enables adoption by stabilizing data protection, contractual delegation and safety-related AI use. Assessments vary across SCOR-DS stakeholder groups, while country-based differences are negligible. This study offers a SCOR-DS-wide, Delphi-based analysis of how AI-driven SCM performance enhancement is expected to unfold by 2040 as a layered, process-uneven configuration rather than a single linear trend. Through a complex adaptive systems lens, it shows that these trajectories operate through distinct mechanisms: schema and network-connectivity changes, shifts in agent composition and self-organization and governance-based boundary setting, thereby advancing theorizing on AI as an adaptive agent in supply networks.