Introduction
Sustainability has become a central concern across industries, but greenwashing—the misrepresentation or exaggeration of sustainability achievements—remains widespread.
Organisations greenwash because genuine improvements are costly, sustainability outcomes are difficult to measure, and audiences reward “green” claims that are hard to independently verify. Under such information asymmetry, overstating achievements can temporarily boost reputation and market share, making greenwashing a rational but socially harmful strategy.
Attitudes toward greenwashing are also heterogeneous. Some actors benefit privately from overstated claims, while others bear reputational or financial risks if misconduct is exposed. This divergence creates persistent tension: without credible auditing, sustainability disclosure often reflects strategic communication rather than verifiable performance.
The emergence of AI-powered auditing and verification technologies offers new opportunities to address this challenge.
Advances in machine learning, natural language processing, and remote sensing enable the analysis of large, diverse datasets—from corporate reports and media content to satellite imagery and sensor data—to detect inconsistencies between claims and observable reality. AI systems can improve the credibility of sustainability information, but adoption is not straightforward. Organizations must weigh the benefits of credibility and trust against costs, implementation challenges, and the possibility of exposing their own shortcomings.
These dynamics highlight the need for research that integrates computational approaches to detection with models of strategic incentives. Building on the greenwashing literature (e.g., Zhang, 2024) and work on cooperative responsibility under externalities (Fang & Cho, 2020), this project will investigate how AI can reshape the landscape of sustainability disclosure. Specifically, we will ask: Under what conditions will organizations voluntarily adopt AI-based auditing and verification tools? To what extent can AI create credible separation between genuine and misleading claims, or do strategic behaviors persist that undermine its effectiveness? How do shared platforms, certification consortia, or cooperative mechanisms affect incentives to adopt AI-based auditing? Which forms of policy intervention best align AI adoption with broader social welfare objectives?
Methodolgy
Methodologically, we will combine game-theoretic and analytical modeling of strategic behavior with computational experiments using AI tools for sustainability auditing. This dual approach will allow us to examine both the incentives surrounding adoption and the technical feasibility of detecting greenwashing across different contexts.
References:
Fang, X., & Cho, S. H. (2020). Cooperative approaches to managing social responsibility in a market with externalities. Manufacturing & Service Operations Management, 22(6), 1215-1233.
Zhang, D. (2024). The pathway to curb greenwashing in sustainable growth: The role of artificial intelligence. Energy Economics, 133, 107562.