AI and Robotic Oversight in ESG Compliance: An Agent-Based Model of Greenwashing

This paper examines how artificial intelligence (AI), agentic systems, and robotics reshape the dynamics of corporate greenwashing, investor trust, and ESG governance. Traditional ESG regimes rely on voluntary disclosure, fragmented auditing, and sporadic regulatory intervention, leaving ample scope for opportunistic signaling. The rise of AI introduces a dual transformation: generative and agentic systems enable firms to manufacture increasingly sophisticated and adaptive forms of greenwashing, while AI-based auditing, robotics, and automated regulation promise unprecedented oversight capacity. To explore these dynamics, we develop an agent-based model in which corporates, investors, auditors, disinformation agents, and AI regulators interact under four institutional scenarios: Baseline, Strong Auditors, Robots on Patrol, and Coordinated Crackdown. The results reveal that under weak oversight, greenwashing proliferates but destabilizes markets, producing cycles of capital misallocation and collapse. Stronger auditors and robotics delay collapse but fail to eliminate systemic fragility, as deceptive firms exploit informational gaps. Only the Coordinated Crackdown scenario, where AI auditors, robotics, and regulators act in concert, sustains investor trust and reduces the variance of capital-allocation outcomes — though the economic and institutional costs of such a regime, which the model does not endogenize, remain a direction for future cost-benefit analysis. These findings advance theoretical debates on ESG governance by demonstrating that systemic credibility depends less on incremental improvements to individual oversight mechanisms than on their integration into coordinated, technology-enabled regimes. The study contributes methodologically by applying agent-based modeling to ESG finance, offering a transparent simulation of how greenwashing and enforcement co-evolve in the age of AI.