Quanta Explains How Pricing Algorithms Can Drive Up Prices Without Collusion

burny_tech · x · 2026-09-23

Quanta Magazine examines the game theory of algorithmic pricing: from Uber rides to airline tickets, prices are increasingly set by learning algorithms that adjust in response to market data. Traditional antitrust enforcement relies on proving explicit collusion, but as UPenn computer scientist Aaron Roth puts it, "the algorithms definitely are not having drinks with each other."

Drawing on a widely cited 2019 paper and follow-up work, the article shows that even simple pricing algorithms—far simpler than deep learning—can emergently arrive at supra-competitive prices without any explicit agreement, posing a fundamental challenge to existing regulatory frameworks.

Original post →

More from AGI Musings

AGI Musings channel →