One bad algorithm, all bad algorithms? Examining public responses to AI-managed CSR practices following AI-driven crises

Wu, L., Zifei Fay Chen, & Tao, W. (2026, June). “One bad algorithm, all bad algorithms? Examining public responses to AI-managed CSR practices following AI-driven crises.” Paper presented at the 33rd International Public Relations Research Symposium, Bled, Slovenia.Abstract: In recent years, organizations have increasingly deployed AI to manage and support CSR activities, from automating grantmaking to forecasting CSR outcomes through agentic AI. At the same time, the growing reliance on AI has introduced new organizational risks. AI failures, ranging from biased algorithmic outputs to safety-related system breakdowns can undermine public trust and cause reputational harms. To this end, a critical question emerges: How do publics respond to an organization’s AI-managed CSR practices when the organization encounters an AI-driven crisis? Situated at the intersection between the increasing adoption of AI in CSR practices and the heightened risks associated with AI failures, this study examines how publics respond to AI-managed CSR practices when AI plays a critical role in both an organization’s CSR practices and the cause of an organization’s crisis. Importantly, we also examine how people’s comfort level with intelligent technologies act as a boundary condition.