Personalization, Disclosure, and Investment Distortions

This paper studies how pricing structures shape AI investment incentives in datadriven markets. We develop a Hotelling model in which effective mismatch costs are jointly determined by firm investment and consumer disclosure. Firms compete either under symmetric personalized pricing or under an asymmetric structure in which one firm personalizes while its rival sets a uniform price. Under symmetric personalized pricing, allocation follows effective costs, and investment affects welfare only through technological improvements. Private and socially second-best incentives therefore coincide. Under asymmetric pricing, however, allocation is governed by total prices. Investment shifts the market boundary and induces strategic price responses by the uniform-price rival, generating a distortion wedge between private and social incentives. The direction of this wedge determines whether investment is socially insufficient or excessive. Our findings show that the welfare consequences of AI investment depend fundamentally on the pricing structure through which competition operates. JEL Classification: L13, L86, D82, O33