What happens when one of the companies reshaping the labor market decides to fund research into how it's reshaping the labor market? That's the situation with OpenAI's newly announced Economic Research Exchange, a program now open to applications from economists and researchers studying AI's effects on jobs, wages, and productivity.
Selected researchers will get access to OpenAI data and resources. The company says it wants credible, independent findings - specifics about which datasets will be made available and how many projects will be funded aren't detailed in the launch announcement.
The underlying research questions are genuinely unsettled. Studies from the past two years have produced contradictory conclusions: some show productivity gains of 30-40% for knowledge workers using AI tools, while others document wage compression in freelance writing, coding, and customer service markets. Both findings can coexist; the real question is which effect dominates, for whom, and over what timeframe.
The conflict of interest is obvious. OpenAI's commercial success depends on AI adoption growing, which makes research that paints AI as net-positive for workers valuable to them. That doesn't automatically invalidate the work - industry-funded studies produce legitimate findings regularly - but publication terms matter. If researchers can only publish with OpenAI approval, or if inconvenient findings quietly disappear, the exchange produces marketing material rather than knowledge.
For people using AI tools daily - marketers who've seen content budgets shrink, developers whose consulting rates have softened, freelancers competing with AI-assisted agencies - answers to these questions have direct practical relevance. What skills remain premium? Which industries are absorbing displaced workers? The Exchange could produce useful data here. Whether it does depends on which researchers are selected and how much independence they're actually given.