Work Autonomy in the Age of AI

• In Event: AI and the Value of Work

Abstract

How generative AI will transform work and workplaces is widely contested, with some foreseeing complete transformations across work sectors and others doubting the efficacy and utility of generative AI tools to complete tasks reliably, ethically, and safely. Scholars have assessed the impacts of AI on a range of fields, from coding to policing, and how different national and social contexts relate to AI adoption at work. Yet few have demonstrated the importance of occupational position and autonomy for determining how AI is used and controlled within workplaces. Drawing on a large, comparative two wave survey fielded in the United States and France and nearly 200 follow up interviews with survey participants , we demonstrate that work autonomy, or the control one has over their own labor, plays a crucial role in AI adoption, control, and interpretation. Leveraging the comparative aspect of this work, we identify consistent trends across advanced capitalist countries as well as how national occupational structures play a crucial role in stratifying access to AI. We also demonstrate how autonomy, and not just occupational position, is deeply entwined with how AI is used, with those in high-autonomy positions, across all occupational positions, adopting AI more frequently than low autonomy workers. High autonomy workers also express a greater sense of control in determining how, when, and if they use generative AI. Combing our survey and interview data, we reveal that low autonomy, low position workers have less access to AI and less control when it appears in workplaces, a set of conditions that can cause resistance among workers. Across workplaces and countries, variations in position and autonomy, or differential authority, plays a key role in the adoption of AI at work.

Authors

• Isabelle Langrock, Middlebury College

• Lauren Clingan, Sciences Po

• Jen Schradie, Sciences Po - Paris