Ethan Mollick says AI agent productivity studies are lacking and may miss large effects

Wharton professor Ethan Mollick, who ran early randomized controlled trials on the productivity effects of AI chatbots, says there has been little comparable research since AI agents emerged last fall.

He attributes the gap partly to the technology being new and partly to difficulties in research design, but he suspects that significant effects may be going unmeasured. This is Mollick's own view as a researcher; the report does not include new study findings.