Urging Congress to focus on the human side of AI integration in the workplace

As artificial intelligence (AI) continues to permeate the American economy, its success will be defined not by the sophistication of its code, but by the psychological readiness of the people using it.

In a recent testimony submitted to the U.S. House of Representatives Committee on Education & Workforce, Subcommittee on Workforce Protections (PDF, 275KB)opens in new window, Dennis P. Stolle, PhD, executive lead psychologist for applied psychology at the American Psychological Association, emphasized that AI is fundamentally a human system rather than a purely technical one. Stolle argued that the economic impact of AI depends heavily on how workers understand, trust, and adapt to these new tools.

Stolle highlighted that AI does not simply happen to a workforce; it is introduced through human decisions that reshape the very nature of work. He cautioned that AI often introduces "hidden labor," where workers spend significant time verifying, correcting, and supervising AI outputs—tasks that are rarely captured in traditional productivity measures.

"AI functions as a job design technology, not just a productivity tool," Stolle noted. Whether AI enhances human capability or creates friction depends on how tasks and responsibilities are allocated between humans and machines.

Central to successful AI integration is the concept of calibrated trust—the ability of a worker to know exactly when to rely on a system and when to question it. Stolle identified several psychological factors that determine this balance:

• Trust: Willingness to rely on AI is driven by perceived reliability and transparency.
• Cognitive load: While AI can simplify complex tasks, it can also increase mental strain if it requires continuous, high-stakes monitoring.
• Fairness and control: Workers are deeply affected by how AI is used in evaluations, scheduling, and task assignments. Opaque systems can undermine organizational commitment.

In its testimony, APA Services called on policymakers to move beyond a narrow focus on technological capabilities and instead address the human and organizational conditions that shape AI use. Stolle’s testimony recommended that future policies should:

• Mandate transparency: Ensure workers understand how AI systems make decisions.
• Support job design: Encourage structures that maintain worker autonomy and fair task distribution.
• Integrate psychological science: Embed psychological competencies and evidence-based evaluation throughout the AI lifecycle, from design to deployment.

"The future of work with AI will be defined by how effectively we understand and manage the human side of that integration," Stolle concluded. "APA and its member scientists stand ready to help ensure AI is developed in ways that promote human flourishing."

For more information, email Corbin Evans.