Generative AI at work: who really benefits?
When a team adopts a generative AI assistant, does everyone improve at the same pace? One of the few large-scale field studies offers a first answer.
The reference study
Erik Brynjolfsson, Danielle Li and Lindsey Raymond (2025, Quarterly Journal of Economics) analyse the staggered rollout of a conversational assistant among customer-support agents. Productivity rises on average, with much larger gains for less experienced agents.
Why it matters
The staggered rollout allows a more robust comparison than a survey. The tool appears to spread top agents' practices to novices; the most experienced gain little.
Limits to keep in mind
One firm, one job, one tool. Nothing guarantees the same pattern in more ambiguous tasks. Acceptance itself is still best described by Fred Davis (1989, MIS Quarterly): perceived usefulness and ease of use. (our executive and employee training programmes)
The limit of common practice
Counting activated licences says nothing about effects on work or their distribution. Adoption can be massive yet benefit only part of the team.
A practical check in Geneva
Before and three months after rollout, measure one simple business indicator separately for junior and senior staff. If the gap does not narrow, the diffusion effect seen in the study has not occurred. To go further: explore the Leading AI Adoption in Your Team training in Geneva, or browse our executive and employee training programmes in Switzerland.
In pictures: Leading AI Adoption in Your Team in Geneva
- leading ai adoption in your team
- Geneva
- research
- training geneva