AISSQC: a new organizational capability linking AI to resilient, sustainable supply chains — Canadian manufacturers with stronger AI-enabled sustainable quality capabilities report higher supply-chain resilience and better sustainable performance, suggesting AI matters most when embedded in complementary organizational routines rather than as technology alone.

Main Finding

The paper develops and validates a new firm-level capability—AI-enabled Sustainable Supply Chain Quality Capability (AISSQC)—and shows that AI investments generate organizational value mainly when they are embedded in complementary organizational capabilities. AISSQC is a multidimensional higher-order capability (five dimensions, 20-item scale). Empirical tests (exploratory factor analysis and PLS-SEM on two independent samples of Canadian manufacturing firms) find: AI Capability → AISSQC; AISSQC → Supply Chain Resilience (SCR) and Sustainable Performance (SP); AI Capability → SCR (direct); and SCR mediates the AISSQC → SP relationship. Theoretical framing: Resource-Based View and Dynamic Capabilities Theory—AI resources create value through capability development rather than via technology adoption alone.

Key Points

• New construct: AISSQC (reflective–formative higher-order capability) integrates AI, quality management, sustainability, collaboration, continuous improvement, and governance.
• Five complementary first-order dimensions (20 items total):
  • AI-enabled Quality Intelligence (AQI): AI for real-time detection, prediction, and decision support for quality.
  • Sustainable Quality Integration (SQI): integrating environmental/social objectives into AI-supported quality processes.
  • Collaborative Quality Orchestration (CQO): AI-enabled coordination among supply chain partners for quality.
  • Adaptive Quality Improvement (AQI-2 / or similar label in full paper): routines for continuous, data-driven process improvement (paper labels include Adaptive Quality Improvement).
  • Responsible Quality Governance (RQG): governance, ethics, and accountability in AI-enabled quality processes.
• Scale development followed best practices: integrative literature review → semi-structured interviews → modified Delphi → cognitive pretesting → EFA → PLS-SEM confirmatory tests.
• Measurement instrument: 20 items on a 7-point Likert scale (content validated by experts; full item list provided in paper).
• Statistical approach: exploratory factor analysis to identify dimensional structure; Partial Least Squares Structural Equation Modeling (PLS-SEM) to validate the reflective–formative higher-order construct and test structural hypotheses (including mediation).
• Main empirical results:
  • Strong positive association: AI Capability → AISSQC (H1 supported).
  • AISSQC positively influences SCR (H2) and SP (H4) (both supported).
  • AI Capability has a direct positive effect on SCR (H3 supported).
  • SCR positively affects SP (H5 supported).
  • SCR mediates part of the AISSQC → SP effect, indicating capability-driven resilience is a key pathway to sustainable performance.
• Context and scope: Canadian manufacturing firms; focus on firm-level organizational routines (unit of analysis = organization).

Data & Methods

• Research design: Sequential exploratory mixed-method approach for construct development and validation.
  • Phase 1: Integrative literature review to define conceptual domain.
  • Phase 2: Semi-structured interviews with academics and senior managers to identify routines and dimensions.
  • Phase 3: Modified Delphi with experts + cognitive pretesting to refine items.
  • Phase 4: Quantitative validation in two independent samples (Canadian manufacturing firms) using:
    • Exploratory Factor Analysis (EFA) to establish dimensionality.
    • Partial Least Squares Structural Equation Modeling (PLS-SEM) to validate the reflective–formative higher-order model and test structural relationships and mediation.
• Measurement: 20 items across five first-order dimensions, 7-point Likert responses. The instrument is reported as content-validated and psychometrically evaluated.
• Theoretical foundations: Resource-Based View (RBV) and Dynamic Capabilities Theory (DCT) guide hypotheses about how AI resources are transformed into capabilities and outcomes.
• Limitations of methods (as reported/implicit):
  • Empirical tests are cross-sectional survey-based (causal claims should be interpreted with caution).
  • Context limited to Canadian manufacturing—generalizability to other sectors/geographies requires testing.
  • The excerpt does not report sample sizes or detailed fit/validity statistics here (see paper for exact n and model diagnostics).

Implications for AI Economics

• Complementarities and complementarities measurement:
  • Economically, the paper emphasizes complementarity between AI capital and organizational/capability capital (AISSQC). Evaluations of AI investments should account for complementary investments (training, governance, routines) rather than treating AI as standalone capital.
  • The validated AISSQC scale provides an operational measure of capability capital that economists can include as a mediator or interaction term in production-function or performance regressions.
• Returns to AI investment and heterogeneity:
  • Returns to AI are conditional on capability development. Models estimating the productivity or profit effects of AI should include AISSQC (or proxies) to avoid biased estimates of AI’s effect.
  • Expect heterogeneity in returns across firms and sectors driven by differences in AISSQC; this can explain why some adopters reap large gains while others do not.
• Policy and public economics:
  • Policy that subsidizes AI hardware/software purchases may have limited impact unless accompanied by support for capability-building (training, process redesign, governance frameworks). Industrial policies should target complementarities.
  • Programs to build governance and responsible-AI capacity (RQG dimension) can reduce negative externalities (e.g., safety, discrimination, environmental harms) and increase social value from AI deployment in supply chains.
• Resilience and system-level externalities:
  • AISSQC enhances supply chain resilience, which has social value beyond firm-level profits (e.g., fewer shortages, lower systemic risk). Social planners and regulators may want to incentivize capability-building to internalize these externalities.
• Measurement & empirical strategy suggestions for economists:
  • Use the AISSQC scale as a mediator in structural models: AI capital → AISSQC → outcomes (productivity, resilience, sustainability).
  • Test for interaction effects between AI capital and AISSQC components to quantify complementarities (e.g., estimated factor returns on AI conditional on capability scores).
  • Employ panel or quasi-experimental designs (diff-in-diff, instrumental variables) to identify causal effects of capability-building and AI adoption on outcomes such as productivity, resilience, emissions, and employment.
  • Explore sectoral dispersion in AISSQC adoption to study diffusion dynamics and market structure implications (entry/exit, concentration).
• Labor and distributional effects:
  • Capability-driven AI deployment that emphasizes quality, sustainability, and governance may change the skill mix required (higher demand for data-literate managers, quality engineers, governance specialists). Economists should account for reallocation and human-capital complementarities.
• Sustainability and social accounting:
  • Because AISSQC links AI to sustainable performance (economic, environmental, social), incorporating such capability measures into ESG and productivity analyses can help quantify trade-offs and co-benefits of AI investments.
• Future empirical questions opened by the scale:
  • What is the marginal product of AI conditional on AISSQC?
  • What is the optimal sequencing (timing) of investments in AI vs. capability-building?
  • How do public incentives for AI adoption compare in cost-effectiveness to subsidies for capability development?

Limitations and recommended next steps for researchers/economists: - Validate the AISSQC scale across other sectors and countries; collect panel data to assess dynamics and causal effects. - Combine firm survey measures with administrative/transactional data (productivity, quality incidents, emissions) to estimate economic impacts. - Explore endogenous firm decisions to invest in AISSQC and potential selection bias when evaluating AI returns.

Summary: The paper provides a theoretically grounded, empirically validated capability construct (AISSQC) and measurement tool that clarifies how AI investments translate into resilience and sustainable performance. For economists, the central takeaway is that AI's economic returns are mediated by organizational capabilities—quantifying and modeling those complementarities is essential for accurate evaluation and effective policy design.