AI maturity, governance and entrepreneurial success: evidence from incubators and accelerators in leading innovation economies

Purpose This study aims to examine how artificial intelligence (AI) maturity translates into entrepreneurial success in incubators and accelerators by testing startup adaptability as a key mechanism and governance as a contextual boundary condition. Design/methodology/approach Cross-sectional survey data from 101 incubator and accelerator professionals in leading innovation economies were analyzed using partial least squares structural equation modeling to estimate direct, mediation and interaction effects. Findings AI maturity contributes to entrepreneurial success primarily through startup adaptability, while AI usage and knowledge transfer show no consistent direct performance effects. Interaction tests indicate no significant moderation effects for ethical governance or sectoral context in this sample, suggesting that contextual influences may operate through upstream enabling conditions rather than short-term interaction patterns. Research limitations/implications The cross-sectional design limits causal inference about capability development over time. Single-informant, program-level perceptions may introduce common-source bias and reduce sensitivity to founder-level learning dynamics. The modest sample size restricts statistical power for interaction testing, and limited variance in governance and sectoral conditions within leading innovation economies may attenuate detectable moderation effects. The marginal AVE for knowledge transfer suggests construct heterogeneity; future research should disaggregate exposure versus enactment and use behaviorally anchored indicators. Longitudinal, multirespondent and sector-balanced designs are needed to validate boundary conditions. Practical implications Managers of incubators and accelerators should prioritize AI maturity investments in data quality, system integration and analytical skills before expecting measurable performance gains. AI-generated insights yield value when embedded into routines that support disciplined iteration, such as structured selection panels, milestone reviews and mentoring processes. Given the weak role of knowledge transfer, programs should shift from exposure-oriented training toward co-execution models that translate learning into enacted practice. Governance should be implemented as enabling infrastructure (auditability, accountability, transparency) that supports trustworthy routine integration. Social implications Strengthening AI maturity and responsible governance in entrepreneurial support organizations can improve transparency, fairness and accountability in venture selection and resource allocation. When AI insights are routinized into decision-making, incubators and accelerators may better support evidence-based experimentation, reduce informational asymmetries and increase the effectiveness of public and private innovation investments. Emphasizing governance as enabling infrastructure – rather than compliance alone – can enhance stakeholder trust and legitimacy of AI-assisted judgments. Over time, such practices may contribute to more resilient entrepreneurial ecosystems and more inclusive access to high-quality support mechanisms. Originality/value The study advances mechanism-based explanation beyond descriptive tool-focused accounts by integrating institutional theory and dynamic capabilities theory to clarify how AI-enabled routines convert capability foundations into program-level outcomes.