Generative AI communications and firm value: Evidence from digital innovation in tourism, travel, and hospitality firms using deep learning

Abstract

This study examines whether firm-initiated generative artificial intelligence (GenAI) communications are associated with firm value in UK-listed tourism, travel, and hospitality (TTH) firms. The final sample includes 177 firms from 2018 to 2024, comprising 68 firms with verified WeChat official accounts and 109 firms retained with zero-valued communication measures. From 105,429 archived posts, we develop a bilingual annotation scheme and supervised deep-learning classifiers to identify GenAI communication and distinguish implementation-salient disclosures from generic mentions. Our preferred measures are share-normalized, capturing both GenAI communication intensity and implementation emphasis relative to total posting activity. Firm- and year-fixed-effects results show that implementation-weighted GenAI communication is positively associated with Tobin's Q and operating performance. Digital innovation partially mediates this relationship, while industry dynamism and complexity strengthen it, but munificence does not. The study contributes a tourism-specific measure of GenAI capability signaling and clarifies the value relevance of implementation-oriented communication in service-intensive digital contexts.