Listed Nigerian firms that disclose AI use show higher measured financial-reporting transparency, and the transparency boost is larger where corporate governance is stronger; however, findings are correlational and based on disclosure content rather than a causal identification strategy.

Main Finding

The study finds that AI adoption by Nigerian listed firms is positively and significantly associated with financial reporting transparency. Greater AI disclosure intensity also predicts higher transparency, and strong corporate governance (board/audit oversight and audit quality) strengthens the positive AI–transparency relationship. The authors conclude the transparency benefits in this emerging-market setting arise from substantive implementation rather than merely symbolic AI signaling.

Key Points

• Context: Nigerian listed firms (emerging market) where digitalization pressure is rising but governance and institutional capacity vary.
• Conceptualization of transparency: a composite, multidimensional index combining annual-report readability, disclosure completeness, audit-report lag, and reporting timeliness.
• AI measurement: both an AI adoption dummy and an AI disclosure-intensity measure (content analysis of annual/corporate reports).
• Theoretical framing: agency theory (AI can reduce information asymmetry), signaling theory (disclosure may be symbolic), and institutional theory (regulatory and investor pressures).
• Hypotheses: (H1) AI adoption increases transparency; (H2) AI disclosure intensity increases transparency; (H3) corporate governance quality moderates (strengthens) the AI → transparency effect.
• Results: All three hypotheses supported—AI adoption and AI disclosure intensity positively predict the composite transparency index, and corporate governance quality amplifies that effect.

Data & Methods

• Sample: 50 listed Nigerian firms, 2019–2024, yielding 300 firm-year observations (purposeful sampling requiring complete annual/corporate governance disclosures).
• Data sources: secondary archival data — annual reports, corporate governance reports, sustainability reports, and Nigerian Exchange disclosures.
• Measurement:
  • Financial reporting transparency: composite index (readability, disclosure completeness, audit lag, timeliness).
  • AI variables: binary adoption indicator + disclosure intensity from content analysis (frequency/extent of AI-related text).
  • Corporate governance quality: proxy using board effectiveness, audit committee independence, audit quality (from disclosures).
• Empirical approach: ex post facto, correlational design using panel regression models; moderation tests to assess governance interaction effects.
• Methodological contribution: combining panel regression with manual/textual content analysis to separate presence of AI disclosure from intensity/substance.

Implications for AI Economics

• Substantive adoption matters: In an emerging market, measurable transparency gains depend on actual AI integration (controls, processes) rather than superficial mentions. For economic models linking technology adoption to information frictions, this supports mechanisms where technology reduces informational asymmetries only if organizational and governance complementarities exist.
• Role of governance complementarities: Corporate governance strengthens the technology–transparency channel. Economists should model complementary investments (board/audit capacity, internal controls) as necessary moderators of digital technology returns in information production.
• Policy and regulation: Findings motivate disclosure standards that require verifiable, specific information about AI use in reporting and internal controls (to distinguish signaling from substantive adoption). Regulators and standard-setters in EMs should consider guidance on AI auditability, explainability, and audit procedures.
• Investor and audit practice: Investors and auditors should incorporate assessment of AI implementation depth (not just mentions) when evaluating reporting quality. Audit firms may need new procedures to assess AI-driven reporting processes.
• Research directions: Need for causal identification (instrumental variables, natural experiments) to address endogeneity; finer-grained measurement of actual AI systems (type, role in accounting processes); sectoral heterogeneity; and the balance of transparency gains versus new risks (explainability, bias, accountability). Limitations: sample limited to Nigeria, purposeful sampling, and possible measurement bias when using disclosures to infer actual AI use.