Internal Seminar: Andras Niedermayer

Competition and Personalized Pricing in the Age of AI

Artificial intelligence creates new possibilities for firms and digital platforms to use detailed consumer information to personalize prices. We study how an information intermediary can strategically reveal such information to competing firms and thereby shape pricing, competition, and welfare. Using Bayes correlated equilibrium, we characterize outcomes implementable through information disclosure, rather than restricting attention to full or no information. We show that partial revelation and correlated price recommendations can sustain full surplus extraction even under competition, while maximizing industry profit can also require inefficient allocations. These possibilities depend on market structure, including which products come under common ownership through mergers. Motivated by the importance of multiproduct firms in empirical industrial organization, we develop computational methods that build on standard empirical demand estimates, enabling researchers to evaluate information design and personalized-pricing counterfactuals in markets with competing multiproduct firms.

Room A406