Banks reporting AI adoption in Nigeria also report more accurate financial statements, faster audits and less information asymmetry; however, evidence rests on a cross-sectional survey of professionals rather than objective performance data.

Main Finding

The study of AI adoption in Nigerian deposit-money banks finds that adopting AI technologies (machine learning, NLP, RPA, predictive analytics) is associated with: - Significant improvements in financial reporting accuracy. - Significant gains in auditing efficiency (including continuous/whole‑dataset auditing and faster anomaly detection). - A reduction in information asymmetry among stakeholders. Organizational readiness and governance structures are identified as crucial enablers that determine the extent of these benefits.

Key Points

• Research goal: assess effects of AI adoption on (i) financial reporting accuracy, (ii) auditing efficiency, and (iii) information asymmetry in Nigeria’s financial sector.
• Design: survey of accounting professionals, auditors, financial managers, and IT specialists in selected Nigerian deposit-money banks.
• Analysis: Structural Equation Modeling (SEM) used to test relationships; three null hypotheses (no effect of AI on each outcome) were rejected.
• Theoretical framing: Diffusion of Innovations (DOI), Technology Acceptance Model (TAM), and Technology–Organization–Environment (TOE) framework — used to explain adoption drivers (relative advantage, perceived usefulness/ease of use, technological/organizational/environmental contexts).
• Mechanisms proposed: automation reduces manual errors; advanced analytics enable whole‑dataset review and fraud/anomaly detection; real‑time processing improves timeliness of disclosure and monitoring.
• Enablers/barriers: organizational readiness, technical infrastructure, staff skills, governance/data‑management frameworks, and implementation costs.
• Caveats noted by the authors: unequal access to advanced technologies could create new asymmetries; adoption effectiveness depends on data quality and institutional support.
• Policy/recommendation summary: financial institutions should invest in infrastructure, staff training, and data governance; regulators and management should strengthen governance to maximize AI benefits.

Data & Methods

• Data source: primary cross‑sectional survey responses from accounting professionals, auditors, financial managers, and IT specialists at selected Nigerian deposit‑money banks.
• Analytical method: Structural Equation Modeling to estimate relationships between AI adoption (independent variable) and three dependent outcomes (financial reporting accuracy, auditing efficiency, information asymmetry).
• Reported empirical result: AI adoption has statistically significant positive effects on reporting accuracy and audit efficiency and a significant negative effect on information asymmetry.
• Missing or limited methodological details in the paper excerpt: sample size, sampling frame and selection procedure, measurement scales/construct operationalization, reliability/validity statistics, SEM specification and fit indices, control variables, and timing (cross‑section vs. panel).
• Inference limits: with the survey/SEM cross‑sectional design, causal interpretation is limited; potential common‑method bias and self‑reporting could inflate associations; sample restricted to deposit‑money banks in Nigeria limits external generalizability.

Implications for AI Economics

• Market frictions and information efficiency:
  • Reduced information asymmetry can lower adverse selection and agency costs, improving capital allocation and potentially lowering cost of capital for adopting firms.
  • Better disclosure timeliness/accuracy could increase investor confidence and improve market liquidity and price discovery.
• Productivity and costs:
  • Auditing efficiency gains imply lower per‑unit audit costs and faster reporting cycles; firms may reallocate audit resources toward higher‑value (judgmental) tasks.
  • Cost savings may be offset by upfront investments in infrastructure, training, and governance — important for cost‑benefit analysis in developing markets.
• Labor and skills:
  • Demand shifts from routine audit/accounting tasks toward data‑science, model governance, and oversight roles; policy and firm-level reskilling programs are necessary.
  • Short‑term displacement risks for lower‑skilled accounting staff; long‑term complementarities for skilled workers.
• Distributional and access concerns:
  • Unequal access to AI (by firm size, capital, regulatory support) could create new information asymmetries and competitive imbalances across firms and markets.
  • Policy interventions (subsidies, shared infrastructure, standards) could mitigate these distributional effects.
• Regulation and governance:
  • Strong data governance, audit of AI models, algorithmic transparency, and standard setting are needed to ensure reliability, limit model risk, and preserve audit quality.
  • Regulators may need to update audit and disclosure standards to incorporate AI‑enabled continuous auditing and model validation practices.
• Research and policy priorities:
  • Need for causal evidence: quasi‑experimental or panel studies that measure objective outcomes (error rates, restatements, audit time/costs, market reactions).
  • Measure heterogeneity: firm size, ownership, regulatory environment, and IT maturity; cross‑country comparisons would clarify how institutional context conditions impacts.
  • Track dynamic effects: long‑run impacts on audit fees, employment composition, capitalization, and market performance.

Practical takeaway for economists and policymakers: AI adoption in financial reporting/auditing appears promising for improving information quality and reducing frictions, but realizing net social gains requires investment in infrastructure, human capital, governance, and regulatory updates — and careful empirical work to quantify costs, distributional consequences, and causal impacts.