Corporate artificial intelligence (AI) governance is becoming an important organizational capability, yet firm-level evidence on why some companies develop stronger governance arrangements than others remains limited. This study develops a Corporate Artificial Intelligence Governance Index (CAIGI) and examines its antecedents among 74 Brazilian listed companies in the Ibovespa. The index comprises 24 items across six dimensionsgovernance structure, AI risk management, data and model governance, human oversight and accountability, monitoring, audit and controls, and transparency and stakeholder protection-and distinguishes public commitments from observable implementation. Using corporate disclosures available up to September 15, 2026, the study tests six hypotheses concerning AI exposure, board technology expertise, general governance quality, regulatory intensity, international exposure, and firm size. Fractional-logit models are used as the primary specification, complemented by OLS-HC3, two-part models, bootstrap marginal effects, leave-one-out analysis, alternative variable definitions, and dimension-specific regressions. AI exposure is the most robust correlate of CAIGI: a one-unit increase in ln(1 + operational AI use cases) is associated with an average 6.65point increase in CAIGI. Board technology expertise is also robust under the primary measure, with a 10percentage-point increase associated with approximately 4.77 additional CAIGI points, although this relationship becomes less precise under a highly restrictive technical-expertise definition. General governance quality is not supported, while evidence for regulation, international exposure, and firm size is weaker after multivariate controls. The findings position AI governance as an organizational response to technological exposure and highlight board technological capability as an important complement in translating AI use into governance structures.