How stage-based generative artificial intelligence (GAI) assimilation impacts sustainability performance: The role of the technology-organization-environment (TOE) framework

As artificial intelligence (AI) evolves rapidly, Generative AI (GAI) has become a transformative force in business operations and strategic decision-making. While its potential to influence gains in sustainability performance is attracting growing attention, the impact outcomes and underlying mechanisms remain underexplored. To address this gap, grounded in the stage-based Diffusion of Innovation (DOI) theory, this study uses a long-horizon event study to trace changes in sustainability performance around GAI adoption events. Drawing on 150 identified events from 2020 to 2023 among Chinese A-share-listed entities, our analysis reveals a U-shaped trajectory in sustainability performance (measured by ESG scores) during assimilation: performance initially dips in the initiation stage (pre-adoption) but recovers and improves significantly in the implementation stage (post-adoption). Furthermore, using cross-sectional regression within the Technology-Organization-Environment (TOE) framework, we identify technological intensity and executive background as key contingent factors that strengthen the positive effect of GAI assimilation on sustainability performance. The study extends current AI-sustainability research by focusing on the performance benefits and contingencies of GAI assimilation. It provides practical guidance for firms to bridge assimilation gaps and aid tech-sustainability synergies.