Larger boards and more frequent meetings predict higher profitability in Palestine’s listed firms, but AI alone shows no direct profit boost; instead AI reshapes governance effects—heightening the value of gender diversity, board education and independence while diminishing the benefit of frequent meetings, hinting at technological substitution for traditional oversight.

Main Finding

• In a sample of 36 Palestine Exchange (PEX) firms (2017–2024, 288 firm‑year observations), board size and board meeting frequency have positive and significant effects on financial performance (ROA and ROE).
• Board independence, gender diversity, board education, and board financial expertise show no significant direct effect on ROA/ROE.
• AI disclosure (measured by frequency of AI‑related terms in annual reports) has no direct effect on financial performance but meaningfully moderates several board–performance relationships:
  • AI strengthens the positive effect of gender diversity on both ROA and ROE.
  • AI enhances the positive effect of board education on ROA.
  • AI reinforces the link between board independence and ROE.
  • The interaction of AI with meeting frequency negatively affects ROA, suggesting a possible substitution effect where AI can replace certain traditional governance activities.

Key Points

• The paper frames the analysis using agency theory, resource dependence theory, resource‑based view (RBV), and contingency theory (AI as a contextual moderator).
• Main hypotheses: governance mechanisms positively affect financial performance; AI affects performance; AI moderates governance–performance links.
• The study’s originality claim: examining AI as a moderator reshaping the governance–performance nexus in an emerging/fragile governance setting (Palestine).
• Practical takeaway for firms: AI can amplify the benefits of some board attributes (diversity, education, independence) but may substitute for routine governance activities (frequent meetings).
• Policy recommendation: develop regulatory frameworks and guidelines to integrate AI into corporate governance, particularly in fragile institutional environments.

Data & Methods

• Sample: 36 firms listed on the Palestine Exchange with continuous data 2017–2024 (initial universe 48 firms; excluded those listed after 2017 or with incomplete disclosure).
• Observations: 288 firm‑year observations.
• Data sources: annual reports and disclosures posted on the PEX; governance variables from governance sections; financials from financial statements.
• AI measure: content analysis of annual reports using MAXQDA 24 to count frequency of AI/technology‑related keywords (proxy for AI adoption/intensity in disclosure).
• Dependent variables: accounting profitability metrics — ROA (net profit / total assets) and ROE (net profit / total equity).
• Key independent variables (board-level): board size, meeting frequency, board independence, gender diversity, board education level, board financial expertise.
• Estimation approach: panel data analysis using random effects regression to test direct and interaction (moderation) effects.
• Limitations acknowledged in the study (implied/related): AI measured via disclosure frequency (an imperfect proxy for actual AI use); emerging‑market context with limited AI adoption; potential endogeneity and omitted variable concerns (study uses random effects rather than fully addressing causal identification).

Implications for AI Economics

• Complementarity vs. substitution: Empirical evidence that AI can act both as a complement (amplifying the positive impact of board diversity, education, independence) and as a substitute (reducing the marginal value of frequent board meetings). Economic models of AI adoption should allow for both complementarities and substitution effects with existing governance capital.
• Measurement issues: Using textual AI disclosure as a proxy is feasible in low‑data settings but noisy. AI economics work should develop richer, multi‑dimensional measures (actual investments, AI projects, usage intensity, process integration) and validate disclosure proxies against firm activities.
• Heterogeneity matters: Effects vary across board attributes and performance metrics (ROA vs ROE). Future empirical work should explore sectoral heterogeneity, firm size, regulatory environment, and stages of AI maturity.
• Policy and regulatory design: Findings support targeted policy — not only incentives for AI adoption but governance standards for AI use in oversight and reporting, transparency requirements, and capacity building for boards in AI literacy. In fragile or emerging governance regimes, regulation can shape whether AI reinforces or undermines traditional control mechanisms.
• Research design recommendations: To improve causal inference in AI economics, incorporate quasi‑experimental designs (difference‑in‑differences around AI adoption events, instrumental variables, or matched comparisons), richer controls for time‑varying confounders, and dynamic models capturing learning curves and costs.
• Corporate governance strategy: From an economic standpoint, firms should evaluate the ROI of AI investments not only for operations but for governance—prioritizing AI that augments cognitive resources (analysis, forecasting) and selective substitution of routine oversight activities to reduce agency costs efficiently.
• Broader implication for market outcomes: If AI systematically changes the effectiveness of governance mechanisms, this could alter market valuation of governance attributes (e.g., investors may value board diversity more when firms deploy AI), affecting corporate governance markets, executive compensation design, and monitoring policies.

Suggested directions for follow‑up research (concise): - Validate disclosure‑based AI measures with actual investment/use data. - Causal identification of AI’s moderating role (natural experiments, phased rollouts). - Sectoral and institutional comparisons across other emerging markets. - Microstudies on how specific AI tools (analytics, monitoring, automated reporting) interact with distinct governance tasks (risk oversight, compliance, strategy).