She codes, she leads: the tipping point for ethical digital transformation in banking

Purpose The purpose of this study is to examine the nonlinear threshold effect of board and executive gender diversity on ethical digital transformation (EDT) in European banks. While gender diversity is increasingly promoted as a governance mechanism for responsible innovation, its impact on the ethical quality of digital transformation remains ambiguous, particularly below critical levels of representation. This paper identifies the precise tipping points at which female leadership becomes a catalyst for fair, inclusive and accountable digital innovation. Design/methodology/approach Using a panel data set of 224 European banks over the period 2011–2022, the authors construct a novel composite index of EDT based on three dimensions: AI ethics and algorithmic fairness, digital inclusion and financial accessibility and user-centric design and accessibility. The Panel Smooth Transition Regression (PSTR) model is used to estimate endogenous thresholds and capture the nonlinear, regime-dependent relationship between gender diversity and EDT, while controlling for board characteristics (size, independence, digital and sustainability expertise) and bank-specific factors (size, capital, profitability and risk). Findings The results of this study reveal a significant critical mass effect. Board gender diversity enhances EDT only when women represent more than 28.70% of board members; below this threshold, the effect is statistically insignificant. Similarly, executive gender diversity significantly improves EDT only beyond a threshold of 52.10%, underscoring the strategic influence of women in operational leadership roles (e.g. Chief Technology Officer and Chief Digital Officer). These findings support critical mass theory and upper echelons theory, demonstrating that women in executive positions directly shape ethical technology governance. This study also shows that the effect is amplified by board digital expertise and independent oversight. Research limitations/implications The sample is limited to European banks, and endogeneity, particularly reverse causality, cannot be fully ruled out. Future research could use instrumental variables or qualitative methods to explore causal mechanisms. Practical implications The identified thresholds provide empirically grounded benchmarks for policymakers (e.g. ECB and EBA) and financial institutions seeking to embed gender-inclusive governance into digital accountability frameworks such as the AI Act and CSRD. Banks should move beyond symbolic representation and target substantive female leadership in both governance and executive functions to ensure technology-driven accountability. Originality/value To the best of the authors’ knowledge, this study is among the first to apply the PSTR model to examine the nonlinear dynamics of gender diversity in the context of EDT. It introduces a robust, multidimensional EDT index and provides endogenous estimates of critical mass thresholds, bridging gender diversity theory, responsible innovation and sustainable digital governance in the financial sector.