From sensing to transforming: The discursive trajectory of AI and the limited prominence of human-centered engagement themes

Abstract

Background Generative Artificial Intelligence (AI) adoption transformed work structures after 2022, raising psychological safety concerns as efficiency priorities overshadow human-centered employee wellbeing. Objective This longitudinal study investigates how engagement is framed as a corporate communication strategy in Standard & Poor's (S&P) 500 technology firms during early large-scale AI adoption (2022–2024). Methods Using IRaMuTeQ, 150 S&P 500 corporate reports (2022–2024) underwent Reinert's Descending Hierarchical Classification, Correspondence Factorial Analysis, and similarity network analysis, employing chi-square ( χ 2 ) marker filtering. Results Findings reveal a three-stage trajectory mapping onto dynamic capabilities: initial sensing (2022), platform seizing (2023), and process transformation (2024). However, human-centered engagement language (meaningfulness, safety, vigor) remained structurally marginalized, failing to form a stable lexical cluster and framing participation strictly through technical integration. Conclusions Public discourse serves as a proxy for enacting dynamic capabilities. Practically, HR leaders should prospectively audit AI communications to ensure technical efficiency narratives do not displace human-centered and psychological safety registers.