Shifting Work Patterns with Generative AI
Eleanor Wiske Dillon, Sonia Jaffe, Nicole Immorlica, Christopher T. Stanton
TL;DR
This study provides field evidence on how access to an integrated generative AI tool, Copilot, reshapes work patterns for knowledge workers in large firms. Using a randomized, firm-scale rollout across 66 firms and 7,137 workers, the authors track time allocation and activity in Office apps via telemetry, applying difference-in-differences and instrumental variable methods to identify causal effects. The key finding is that Copilot reduces Outlook email time by about 1.4 hours per week (ITT) and roughly 2 hours (LATE) while also creating small opportunities to work outside typical email windows, but it does not trigger broad shifts in task quantities or types, nor measurable changes in Teams meetings or Word document output. The results highlight the importance of firm-specific adoption practices and the need for complementary organizational changes to realize larger productivity gains from AI, suggesting that real-world benefits may accrue gradually through co-invention and workflow evolution rather than immediate, across-the-board productivity jumps.
Abstract
We present evidence from a field experiment across 66 firms and 7,137 knowledge workers. Workers were randomly selected to access a generative AI tool integrated into applications they already used at work for email, meetings, and writing. In the second half of the 6-month experiment, the 80% of treated workers who used this tool spent two fewer hours on email each week and reduced their time working outside of regular hours. Apart from these individual time savings, we do not detect shifts in the quantity or composition of workers' tasks resulting from individual-level AI provision.