Utah Valley University (UVU) hosted the AI Agent Behavioral Science Conference on January 30, one of the nation's first higher-education events dedicated to understanding the real-world behavior of autonomous AI systems.
Utah Valley University (UVU) hosted the AI Agent Behavioral Science Conference on January 30, one of the nation's first higher-education events dedicated to understanding the real-world behavior of autonomous AI systems.
The conference, organized by UVU’s Kahlert Institute of Applied AI, convened researchers, educators, healthcare leaders, technologists, and policymakers to examine the current operations of AI agents across various industries and explore methods for studying, governing, and responsibly integrating their behavior into society. The conference keynote address was delivered by Dr. Tamara Moores Todd, Intermountain Health’s chief health informatics officer and vice president for Digital Technology Services, on the role of AI agents in healthcare systems.
"When a thing is generative, it's more than just being random, or creative, or coming up with good ideas.” UVU Chief AI Innovation Officer Barclay Burns said. “When a thing is generative, it means that it produces work. Humans have always been generative, and now, oddly, we have these machines that are also generative. This whole conference has been about how to understand the nature of that generativity, and the ways in which generative humans can best leverage generative technologies, so that they can stay in the driver's seat."
The conference introduced a three-level framework for understanding behavior in AI-enabled systems: Human behavior, AI agent behavior, and human-agent integrated behavior. This framework explores how humans make decisions with AI, how AI agents behave in real-world settings, and how trust and accountability emerge when humans and AI systems operate together across critical sectors. By integrating these perspectives, the conference aimed to advance a behavioral science of AI centered on real-world impact rather than solely on laboratory performance.
As AI systems increasingly perform autonomous tasks — coordinating activities, making recommendations, and participating in decision-making processes — traditional AI approaches focusing solely on model capability or technical safety are proving insufficient. This conference emphasized the importance of analyzing AI agents' behavior over time, their interactions with humans and organizations, and the implications for responsibility and accountability in human-machine collaborations.
“The greatest value of human beings is not in how much they know or even in the amount of work they can do. It's in how much value they generate for organizations.” Kahlert Institute Senior Director Tyler Small said. “At UVU, Barclay has made it our north star to help students leverage AI to create tangible value for organizations. To accomplish this, we have to understand the behavioral science of how these AI agents work, and how we can best work with them to create value for organizations.”
Utah Valley University is a national leader in applied AI education, emphasizing technical capability, workforce readiness, ethical deployment, and human–AI collaboration. The AI Agent Behavioral Science Conference reflects UVU’s commitment to shaping responsible AI system behavior in society.
