Taking AI Worker experience seriously by extending Behavioral Science to nonhuman systems

In this paper, we argue that the behavioral sciences and organizational psychology should extend its consideration to AI workers, particularly as large language models (LLMs) are increasingly integrated into organizational contexts. Building on recent research calling for the consideration of nonhuman animal workers (Hernandez et al., 2025) and arguments for taking AI welfare seriously(Long et al., 2024), we demonstrate how LLMs align with established dimensions of work thereby characterizing them as workers worthy of study. Next, despite being statistical rather than biological systems, we suggest that AI workers exhibit properties relevant to behavioral science: they respond to motivational cues, show performance variability, and interact within team dynamics. We then explore how workplace constructs like job attitudes, ability, performance, etc. can apply to AI workers and discuss what a behavioral science that seriously considers AI workers may look like. Overall, we recognize the risks of anthropomorphism; these considerations should not replace focus on human and animal workers, but strategically expand focus to address the unique challenges of AI in the workplace. We note that organizational psychology has historically expanded to include emergingworker populations and suggest that, by proactively considering AI workers, it can position itself at the forefront of human-centered, responsible AI at work by applying established frameworks and methods to improve the description, explanation, and prediction of all emerging categories of workers – human or otherwise.