Feedback in the age of AI: understanding user reactions to AI and human agents
Meiling Yin, Mingyeong Jeon · Behaviour and Information Technology · 2026
Interactions between firms and users on online platforms critically shape user attitudes, highlighting the need for strategies aligned with these dynamics. This study examines how feedback valence (positive vs. negative) and feedback provider (AI vs. human) jointly influence user attitudes towards platform firms. Across three behavioural experiments, we find that negative feedback is more readily accepted when provided by AI agents, increasing feedback acceptance and fostering more favourable attitudes towards platform firms. Perceived fairness emerges as the key mechanism: negative feedback from AI rather than human agents elicits higher fairness perceptions, mitigating the negative impact of adverse feedback. Task importance further moderates these effects. AI-provided negative feedback is perceived as fairer when tasks are low in importance, whereas human-provided negative feedback is judged fairer for high-importance tasks. These findings provide theoretical insights and practical guidance for aligning feedback strategies with user perceptions on digital platforms.