Firms reporting AI use in HRM also report better performance and efficiency, largely because AI appears to speed and streamline HR processes; however, evidence comes from a single cross-sectional survey and cannot confirm causation.

Main Finding

AI adoption in HRM is positively associated with both organizational performance and organizational efficiency. Improvements in HR process efficiency partially mediate these relationships, indicating that streamlining HR processes is a key pathway through which AI generates organizational benefits.

Key Points

• Scope and contribution: Empirical study of AI in HRM within IT-related organizations in Lagos State, Nigeria; novel focus on HR process efficiency as a mediator linking AI adoption to both performance and efficiency.
• Theoretical framing: Integrates Technology Acceptance Model (TAM), Resource-Based View (RBV), and Dynamic Capabilities Theory (DCT) to explain adoption, strategic value, and organizational adaptation.
• Mechanisms identified:
  • Direct: AI enables data-driven decision-making (predictive analytics, candidate matching, etc.) that supports better resource allocation and productivity.
  • Mediated: AI improves HR process efficiency (speed, accuracy, cost-effectiveness of recruitment, payroll, performance appraisal), which in turn raises organizational outcomes.
• Results: Path analysis on three hypotheses shows statistically significant positive relationships between AI adoption → HR process efficiency → (organizational performance and efficiency), plus direct positive links from AI adoption to the two outcomes.
• Contextual note: Adds evidence from a developing-country IT sector, addressing a gap in empirical studies on AI in HRM in such settings.

Data & Methods

• Sample: Cross-sectional survey of 452 HR personnel and employees employed in IT-related organizations in Lagos State, South-West Nigeria.
• Measures: Self-reported measures of AI adoption in HRM, HR process efficiency, organizational performance, and organizational efficiency (as conceptualized in the HRM and IS literature).
• Analysis: Path analysis testing three hypotheses, including mediation tests for HR process efficiency.
• Theoretical lenses guiding measurement and interpretation: TAM (perceived usefulness/ease of use and adoption), RBV (AI as a strategic resource combined with complementary capabilities), and DCT (sensing, seizing, transforming to capture value).
• Limitations implied by methods: cross-sectional design (limits causal inference), sector- and location-specific sample (generalizability), and reliance on perceptual/self-reported measures.

Implications for AI Economics

• Productivity and cost structure: Evidence that AI in HRM reduces time/cost per HR transaction and administrative errors suggests measurable micro-level productivity gains. Firms can reallocate HR labor toward higher-value tasks, shifting the composition of labor demand (less routine clerical work, more strategic/analytic roles).
• Returns to AI investment depend on complements: RBV and DCT framing implies that AI capital yields greater economic returns when paired with firm-specific data, human skills, and organizational routines. Economically, this raises the importance of complementary investment (training, process redesign) for realizing productivity gains.
• Labor market effects and skill premium: As HR tasks are automated, demand will shift toward higher-skill HR workers (analytics, strategy), potentially increasing wage premiums for complementary skill sets and widening within-firm skill differentials.
• Adoption heterogeneity across firms/countries: Findings from Nigeria’s IT sector show benefits in developing-country contexts but highlight that realized gains may vary with firms’ dynamic capabilities and institutional factors. Economists should expect heterogeneous diffusion and returns across sectors and countries.
• Aggregate and distributional considerations: Widespread adoption could lower HR-related operational costs across firms, affecting profitability, prices, and potentially employment levels in administrative HR roles. Redistribution effects (task reallocation, retraining needs) and the pace of adjustment matter for labor market outcomes.
• Policy and measurement: Policymakers aiming to capture the social gains should facilitate complementary investments (training, digital infrastructure) and monitor algorithmic fairness and bias risks that can impose social costs. For empirical work, combining perceptual surveys with objective firm-level performance and cost measures (productivity, hiring costs, turnover metrics) will improve economic assessment of AI’s impact.
• Research agenda for AI economics: The mediated link via process efficiency suggests economists should model AI as a capital good whose productivity interacts with organization-specific complementarities and dynamic adjustment costs; use panel data and administrative outcomes to estimate causal returns and general-equilibrium labor effects.

(Study: Akintola et al., "Exploring How the Adoption of Artificial Intelligence in Human Resource Management Enhances Organizational Performance and Efficiency", JBDI 2026, n=452, Lagos IT firms.)