Science favors studies with the most citations. After reviewing the Google Academy database, which contains over 1,000 articles and books published in the last decade on the topic of “AI + Economics / Business / Market,” 257 were selected, each with more than 100 citations (as of August 2025).
The overall impression from these studies is that research on AI in business is mainly conducted in three regions: America (40%), Europe (30%), and China (24%). One reason China's lower number here might be due to the ban on Google database in China.
The quality of study (number of citations) primarily depends on the author, rather than the publication year or the journal’s ranking, despite the presence of many specialized journals. The top three most cited scholars are Daron Acemoglu, a Nobel Prize-winning economist from Massachusetts Institute of Technology, Erik Brynjolfsson from Stanford University, and Ajay K. Agrawal from the University of Toronto.
The selected studies prioritize discussion on four key topics:
- AI's impact on the labor market and productivity;
- AI assistance in decision-making and management;
- AI as a tool in marketing and consumer behavior;
- AI as a driver of economic growth.
For example, the two most cited papers focus on AI in management (Duan et al., 2019, with 3,428 citations) and labor (Acemoglu et al., 2019, with 2,970 citations).
What does science say about AI's effects on business? The beauty of science is that it doesn’t have one answer. Still, some studies show similar results. Among the top ten papers most-cited overall and on average per year, two are synchronized in their outcomes:
- An experiment with 453 professionals found that half of those randomly assigned to use ChatGPT during writing tasks completed their work 40% faster and produced work of 18% higher quality. As a result, inequality among workers decreased (Noy et al., 2023).
- Data from 5,172 customer support agents shows that access to ChatGPT increased their productivity, measured by issues resolved per hour, by an average of 15%. The larger gains (up to 30% rise of productivity) were seen among less experienced and lower-skilled agents, encouraging them to learn English, be more polite, and contact a manager less often (Brynjolfsson et al., 2025).
So, science already knows that ChatGPT boosts productivity. The practical takeaway is that AI isn’t just good and welcomed in business – it’s now a must-have tool.
