Pricing, Mergers, and Regulation Under the AI Flywheel Effect

ABSTRACT This paper constructs a theoretical model to explore the operating strategies of generative AI firms across two markets: one directly using AI products and the other indirectly through AI‐enabled industries. The model examines how pricing, mergers, and regulatory strategies are influenced by the AI flywheel effect, a virtuous cycle where AI product adoption and user data feedback enhance performance and promote further adoption. Our two‐period model reveals that AI firms can boost profits by lowering prices in the direct market to acquire more usage data. However, excessive data collection per user may decrease profits, depending on the size of the indirect market. The AI flywheel effect's positive externalities and incentive issues impact both data collection strategies and pricing and merger decisions. Our research indicates that the pricing and merger decisions of firms crucially depend on the size of both markets and the rate of improvement in AI technology. If Market B is large, mergers may simultaneously reduce consumer surplus in both markets, a situation that may require regulation; whereas if Market B is small, mergers can actually increase consumer surplus, and policymakers can encourage interfirm connections. These findings provide valuable insights for balancing market efficiency and consumer welfare in the AI era.