Tool, or agent? Trulaske researcher examines what factors into our acceptance of AI

When it comes to allowing artificial intelligence (AI) into our lives, the role we view it playing predicts our willingness to accept it, says Bingqing Li, an assistant professor of marketing at the University of Missouri’s Robert J. Trulaske, Sr. College of Business.  

Li and research partners Edward Yuhang Lai and Xin (Shane) Wang examined what drives people to accept or reject AI, especially as it becomes more autonomous and socially present. Their meta-analysis of nearly 120,000 participants discovered that certain AI features, including capability, transparency, role, expertise scope and anthropomorphism, impact acceptance differently depending on whether AI is perceived as a tool to be used or an agent capable of acting on its own.

“As AI becomes increasingly capable of acting autonomously and interacting in socially intelligent ways, it is now viewed not only as a technological tool but also as an agentic entity,” Li pointed out. “However, current understanding of people’s willingness to accept AI has largely focused on the perspective of AI as a tool, leaving gaps in how agentic aspects shape acceptance.”

Li’s innovative study is the first meta-analysis to synthesize human acceptance of AI through a dual-perspective lens (i.e., AI as a tool versus AI as an agent) and to map actionable design features within a User-Centered Design framework. The researchers also launched a web-based tool that lets users explore the data interactively.

“This research addresses a critical question: As AI grows more powerful, agent-like, and embedded in daily life and work, what makes people comfortable accepting it and how can businesses design AI systems that people are willing to use?” Li asked.

The findings have implications for product development and marketing teams, AI and UX designers, tech media, industry practitioners that are responsible for designing or deploying AI systems, public sector stakeholders involved in AI adoption, interdisciplinary researchers exploring human-AI interaction and others. 

For those looking to use the research to increase user acceptance of AI, Li offered the following takeaways:

1. Know the perspective: Which AI features matter most depends on whether people perceive and accept the AI as a tool or as an agent.
2. Design for acceptance: Develop AI features (e.g., capability, expertise scope, role) that increase people’s willingness to use AI.
3. Apply User-Centered Design: Tailor AI design and messaging using the provided User-Centered Design roadmap to match user needs and context.

Mizzou’s Trulaske College of Business is committed to working to advance transformative technologies as one of its three pillars of excellence. 

“As AI technologies like ChatGPT and autonomous agents enter mainstream use, businesses and policymakers urgently need evidence-based guidance on how to encourage user trust and adoption in a rapidly evolving technological landscape,” Li said.

Li’s study, “From Tools to Agents: Meta-Analytic Insights into Human Acceptance of AI” was published in the June 2025 Journal of Marketing.

Mizzou’s Robert J. Trulaske, Sr. College of Business prepares students for success as global citizens, business leaders, scholars, innovators and entrepreneurs by providing access to transformative technologies, offering experience-centered learning opportunities and fostering an entrepreneurial mindset.