When Dynamic Competition and Static Competition Merge: Antitrust, Innovation Questions, and the Case of Generative Artificial Intelligence

Salil K. Mehra October 23, 2025

Antitrust Law Journal View Original Source

• 
  Back
  to Top

Summary

• For much of the last half century, static efficiency has been posited to be in tension with dynamic efficiency. Theorists such as Schumpeter and Williamson introduced models suggesting that firms might compete to innovate to produce new and better products and/or to produce more output at less cost—and that the incentive to do so might conflict with antitrust policies aimed at improving static competition.
• The conflict paradigm of static versus dynamic (and productive) competition is no longer reliable as a guide to better antitrust enforcement and policy. Frequently, dynamic competition has been invoked as cudgel to beat back conventional antitrust enforcement with claims that a defendant firm or its industry is “innovative.”
• At times, this innovation-based argument is deployed to magnify another assertion: that false positives in antitrust enforcement are more harmful than false negatives, because the latter is claimed to be more self-correcting than the former.
• Deciding not to pursue antitrust enforcement due to forecasted innovation means making a tradeoff between the welfare of current versus future consumers. But if competitive markets yield more innovation than monopolistic ones, then no such tradeoff is needed.
• Deeming particular business models, firms, or industries to be innovative, and thus relatively shielded from antitrust enforcement, does not fit with established antitrust law and practice.

Read the full piece here.