This paper examines whether companies adopting artificial intelligence (AI) demonstrate stronger alignment between what they disclose in ESG reports and their actual ESG performance, a key issue in debates around digital greenwashing. Using a hybrid method that combines rule-based filters with zero-shot transformer models, AI adopters were identified within the S&P 500 index. A post-double LASSO variable selection process was applied to estimate the causal impact of AI adoption on ESG transparency, followed by several matching and reweighting techniques, including nearest neighbour matching, entropy balancing, and covariate balancing propensity scores (CBPS). Before matching, clear imbalances in industry representation and financial profiles were observed; these were substantially reduced after adjustment. Across all estimation approaches, results consistently showed that AI-adopting firms have significantly higher alignment between ESG disclosure and performance, with average treatment effects ranging from 0.08 to 0.17 standard deviations. These findings suggest that, contrary to concerns about strategic disclosure, firms that integrate AI may show greater reporting discipline. The results have important implications for ESG evaluation standards, regulatory policy, and the development of trustworthy AI governance frameworks.