2026 · WP
Augmentation or substitution? Field evidence on generative AI adoption in knowledge workWorking Paper
Working paper · Cornell ILR
The JEM Lab for Generative AI at Work, led by John E. McCarthy at Cornell's ILR School, studies the human side of generative AI in the workplace — how new tools change tasks and skills, who captures the gains, and how organizations can deploy AI in ways that empower rather than displace workers.
The JEM Lab — short for the lab of John Edward McCarthy — is housed at Cornell University's School of Industrial and Labor Relations. We study how generative AI is being adopted inside real organizations and what it means for the workers, managers, and institutions that surround it.
Our methodological roots are in the empirical study of work: organizational ethnography, social network analysis, longitudinal field studies, and partnerships with firms, unions, and public-sector employers. We bring those tools to bear on a moving target — a technology whose effects depend enormously on how, by whom, and under what governance it is deployed.
We believe that the most useful research on AI and work is conducted with workplaces, not just about them. Most of our projects are field studies built on long-running partnerships, and our results return to the practitioners who made them possible.
Field studies of how knowledge workers actually use generative AI day to day, and how adoption reshapes tasks, skills, and discretion.
How institutions for employee voice — from labor-management partnerships to AI governance committees — affect what gets built and who benefits.
Tracing how knowledge, norms, and practices diffuse through workplaces, and how AI tools alter those networks of collaboration.
Translating research into policy: gain-sharing, training, and governance frameworks that make AI adoption work for workers.
A longitudinal panel of knowledge workers tracking generative AI tool adoption, task changes, and well-being. Anonymized waves released annually for academic use.
A validated measure of managerial openness to employee voice, used in our ILR Review work. Free for academic and non-commercial use under CC-BY.
Replication package for McCarthy & Levin (2019, JAP). R and Stata code, simulated data, and pre-registered analyses for the dormant ties study.
A growing collection of teaching cases on generative AI in HR and labor relations, used in ILR coursework and the eCornell AI & the Future of HR certificate.
John is an associate professor in the Department of Global Labor and Work at Cornell University's ILR School. His research examines how to build and sustain collaborative organizations, the impact of employee participation on workers and broader organizational outcomes, and — increasingly — the transformative effects of generative AI on the future of work.
Before joining Cornell in 2015, John was a postdoctoral fellow at MIT's Sloan School of Management (with Tom Kochan) and a visiting doctoral student at The Wharton School (with Matthew Bidwell). He received his PhD from the School of Management and Labor Relations at Rutgers University in 2014. His work has appeared in ILR Review, Industrial Relations, British Journal of Industrial Relations, Journal of Applied Psychology, and Personnel Psychology. In 2020 he received the John T. Dunlop Outstanding Scholar Award.
Studies generative AI adoption among professional services workers.
Working on the GenAI@Work Panel; interests in worker voice and AI governance committees.
Studies social network effects of AI tool diffusion within firms; mixed methods, two field sites.
Investigating gain-sharing arrangements in early enterprise GenAI deployments in the public sector.
Practitioner case studies on generative AI rollouts in unionized environments.
Survey design and panel maintenance; interested in AI governance and HR analytics.
Project manager for field engagements and the AI & HR teaching case library.
ILR undergraduate; interview coding, literature review, and panel data cleaning.
ILR undergraduate; field-site logistics and qualitative data management.
We collaborate with firms, unions, public-sector employers, and other researchers who want to study generative AI in real workplaces — carefully, and over time. We also advise PhD and Master's students at Cornell ILR. If your work overlaps with ours, we'd love to hear from you.