AI Disclosure Intensity and Short-Term Stock Returns Evidence from Spring 2026 Earnings Calls

Abstract

This study explores whether the intensity of AI (artificial intelligence) related language used during the Spring 2026 earnings calls is associated with short-term abnormal stock returns and whether this relationship differs between technology-oriented and comparison firms. Using a cross-sectional event study framework, this study analyzes a sample of 50 publicly traded large-cap U.S. firms, consisting of 25 technology-oriented firms and 25 comparison firms from non-technology industries. AI disclosure intensity was measured as the proportion of AI-related terminology relative to total transcript word count. In contrast, market reactions were measured using cumulative abnormal returns (CAR) over a three-day [-1,+1] event window surrounding each earnings announcement. The results indicated that technology-oriented firms exhibited higher average AI disclosure intensity and greater dispersion in cumulative abnormal returns than firms in the comparison cohort. Regression analysis suggested that the association between AI disclosure intensity and short-term abnormal returns varies across industry groups, with AI-related communication appearing to have a stronger relationship with market reactions, particularly among technology-oriented firms. However, these findings should be interpreted as statistical associations rather than any evidence of causality. Moreover, other earnings-related information released simultaneously may also influence stock-price movements. Overall, the study provides exploratory evidence that investors may evaluate AI-related corporate disclosures differently depending on industry context. These findings are consistent with theories of corporate signaling and information credibility, although additional research using larger samples, longer time periods, and more comprehensive control variables is required to establish the underlying mechanisms.