Cal State Fullerton study shows employees work best with AI as option, not mandate
A Cal State Fullerton researcher studying human-AI collaboration in the workplace is pushing back against the assumption that more AI automation equals better business outcomes. The answer, according to Hunter Phoenix Van Wagoner, an assistant professor of management at CSUF, depends heavily on how companies implement the technology.
Van Wagoner examined how 1,100 participants responded to AI recommendations in high-uncertainty tasks, specifically facial recognition and hiring decisions. His 2025 research, published in the Journal of Organizational Behavior, found that employees align with AI suggestions most effectively when they view the technology as a tool they can choose to use-not something imposed on them.
"If employees feel compelled to follow AI recommendations, perhaps they won't use it," Van Wagoner said.
The employee perspective matters
Ninety percent of businesses now use some form of AI, according to McKinsey & Company's 2025 report. Yet most organizations remain in early stages of deployment. Sixty-two percent are experimenting with AI agents that can act autonomously without constant human oversight.
Van Wagoner's concern centers on what gets lost in this rush toward automation. Employees fear being "fired by algorithm." They bring psychology, ego, and what researchers call "inclusion behavior" to their work-factors that make human decision-making unpredictable but also valuable.
"We're messy," Van Wagoner said, acknowledging the human element that doesn't fit neatly into efficiency calculations.
Training is the critical variable
The best outcomes emerge when managers train employees to use AI as a collaborative tool while preserving their ability to think critically. This balance-maintaining human autonomy while benefiting from AI recommendations-represents what Van Wagoner calls "the sweet spot."
The challenge is sustained. As companies scale AI across operations, they must resist the temptation to treat algorithms as final decision-makers. The alternative-employees who disengage from tools they don't trust or feel forced to use-undermines the entire premise of human-AI teamwork.
For management professionals implementing AI systems, the research suggests a straightforward principle: treat AI as a recommendation engine, not a replacement for human judgment. The technology works best when employees retain the authority to accept, reject, or modify its suggestions.
Learn more about AI for Management or explore how AI for Human Resources can support better hiring practices while maintaining human oversight.
Your membership also unlocks:
