Using customized, conversational AI agents in leadership and management research: Benefits, practical illustrations, and best practices

Conversational AI agents—systems capable of holding intelligent conversations with human users—are rapidly reshaping how organizations operate, from leadership development and employee training to internal communication. Consequently, researchers across leadership, management, and the broader social sciences are beginning to examine how these agents affect organizational processes, employees, and workplace outcomes. Yet, existing studies still often rely on scenario-based methods that—while offering experimental control—are limited in ecological validity. Recent advances in no-code platforms mark a turning point: researchers can now design and deploy customized, conversational AI agents without requiring any technical expertise. This development makes it more feasible to conduct empirical studies based on real-time, interactive experiences with functional AI agents rather than imagined scenarios. These agents can represent a variety of organizational actors, including leaders, coworkers, or subordinates; display diverse characteristics and behaviors; and be implemented in complex study designs across lab and field, experimental and observational, and both quantitative and qualitative methodologies. We demonstrate the power of this approach through three empirical studies (N = 789), showing how interactions with customized, conversational AI agents can meaningfully shape participants’ perceptions, attitudes, and behaviors in incentivized settings. Introducing a novel, open-source tool called ResearchChatAI as an illustrative example, we outline how such studies can be designed and deployed—and critically reflect on the practical and methodological trade-offs involved. We showcase how such tools enrich the methodological toolkit of scholars and pave the way for more valid, realistic, and scalable leadership and management research on as well as with AI.