AI Partnerships and Competition: Damned if You Buy, Damned if You Don’t
Abstract
In this paper, we examine the rise of strategic partnerships between large technology firms and AI startups, arguing that these arrangements—often structured to avoid traditional merger-control thresholds—have sparked undue antitrust concern despite limited evidence of competitive harm. Situating these “AI partnerships” within the broader framework of competition policy, we contend that such collaborations are frequently procompetitive, enabling startups to access capital, infrastructure, and distribution channels necessary to scale innovation in a rapidly evolving market. While regulators have raised theoretical concerns about foreclosure, switching costs, and concentration, empirical evidence to date suggests that AI markets remain dynamic, with robust entry, falling prices, and intensifying rivalry across the ecosystem. We further argue that expanding merger-control rules to capture non-controlling partnerships risks undermining legal certainty and deterring efficiency-enhancing collaborations. Ultimately, we conclude that AI partnerships are more likely to promote, rather than hinder, competition and innovation, cautioning against regulatory overreach that could stifle the development of emerging AI markets.