The FTC and State Case Against Amazon Highlights Risks and Impacts from Using Pricing Algorithms

Background

In the fall of 2023, the FTC and 17 (now 19) states and territories sued Amazon.com (Amazon) for violating federal and state antitrust laws.  The enforcers’ complaint alleges that Amazon violated Section 2 of the Sherman Act by unlawfully monopolizing two interrelated markets, the online superstore market and the online seller services market.  The complaint alleges that Amazon excluded competitors and harmed consumers through Amazon’s use of two pricing algorithms and imposing a tie between Amazon’s Prime customers and Amazon’s fulfillment service.  The complaint further alleges that Amazon violated Section 5 of the FTC Act and parallel state laws by using a third, secret pricing algorithm called “Project Nessie.”  Finally, the states and territories brought additional similar claims under their respective state and territorial laws.  Amazon moved to dismiss the complaint in its entirety, but on October 7, 2024, the district court unsealed an opinion denying Amazon’s motion to dismiss the federal antitrust claims (i.e., the Section 2 and Section 5 claims), though some state claims were dismissed without prejudice.

Amazon’s Three Challenged Algorithms

The details of Amazon’s three pricing algorithms underscore the FTC and states’ current aggressive stance on antitrust enforcement and the likelihood that Amazon’s conduct harmed numerous clients. 

All three algorithms rely on Amazon’s extensive surveillance network, which detects prices offered on similar goods from across the internet. The enforcers plausibly alleged that the three algorithms use this surveillance data to harm competition in different ways. 

The first “anti-discounting” algorithm applies when a third-party seller is offering goods on Amazon. In this scenario, Amazon allegedly excludes third-party sellers from the all-important Buy Box if Amazon detects the same good is priced lower off Amazon.  Accessing the Buy Box is allegedly critical for third-party sellers, as this is where 98% of sales on Amazon occur.  Amazon’s first algorithm thus allegedly punishes third parties for offering lower prices off Amazon, which harms shoppers because Amazon’s fees are allegedly much higher than fees charged by other platforms.  The district court, following the logic from its earlier rulings on similar allegations brought by private plaintiffs, held that the government enforcers here plausibly alleged that this anti-discounting algorithm excludes competition to maintain Amazon’s monopolies.

Amazon’s second “anti-discounting” algorithm applies when Amazon is offering goods it owns for sale, i.e., when Amazon functions as a traditional retailer and sells products to shoppers directly.  This algorithm relentlessly copies price changes Amazon detects elsewhere, whether the prices go up or down.  The complaint plausibly alleges that Amazon implemented this price-copying algorithm with the intent to avoid a “perfectly competitive” market where rivals continually lower prices by competing away profits, to the benefit of shoppers.  The district court held that the enforcers’ complaint plausibly alleges that Amazon violates Section 2 by using this price-copying algorithm.

Amazon’s third algorithm is not an anti-discounting one, but rather an unfair method of competition called “Project Nessie.”  Project Nessie predicts whether other companies will match Amazon’s price hikes, induces those companies to raise their prices to Amazon’s new price, and then overcharges shoppers both on and off Amazon.  The enforcers allege that Amazon turned Project Nessie on and off during periods of increased media focus on Amazon’s prices.  The district court ruled that the enforcers plausibly alleged that Amazon’s use of Project Nessie violated Section 5 of the FTC and parallel state laws.

Takeaways

The district court’s ruling on Amazon’s first pricing algorithm, which penalizes third-party sellers for offering lower prices off Amazon, demonstrates the risks associated with any algorithm that makes pricing decisions based on prices detected on another company’s website or app.  While Amazon claimed that it used this algorithm to display low prices to shoppers, the district court held that it was not proper to consider this justification at the motion to dismiss stage.  Thus, Amazon will have to endure the cost of discovery and summary judgment before its side of the story can be told.  Clients using algorithms that consider similar “off-platform” data should carefully assess the risks and costs of potential antitrust litigation against the benefits a pricing algorithm produces.

Next, the district court’s ruling on Amazon’s second pricing practice, the price-copying algorithm, takes an expansive view of how the FTC and states could bring a monopolization claim.  This is the first time any court has held that the regular practice of monitoring and copying rivals’ prices, without any claim of collusion, can violate the Sherman Act.  Notably, the district court was persuaded by the enforcers’ well-pleaded allegations of the anticompetitive intent behind Amazon’s use of this price-copying algorithm, which was to undercut the basic assumptions of classical microeconomics that competition will reduce prices as profits are competed away.  Clients should analyze whether they (or their vendors) use similar price-copying algorithms, and whether there is similar evidence of anticompetitive intent in their own documents.

The district court’s ruling on Amazon’s third algorithm, Project Nessie, is notable for being the first “standalone” Section 5 claim upheld in federal court in over 40 years.  Most competition claims brought under Section 5 of the FTC Act incorporate claims of anticompetitive conduct under the Sherman Act or other antitrust laws.  The FTC’s allegations against Project Nessie, however, plausibly stated a claim for an unfair method of competition under Section 5 by itself, without any reference to another antitrust law.  While the FTC had previously brought such standalone claims in the first half of the 20^thcentury, the FTC had not successfully done so since at least the 1980s.  The district court’s ruling on Project Nessie is also important because it was reached without a discussion of Amazon’s market power and without any form of direct communication or agreement between market actors on price, which normally forms the basis for a price-fixing claim. Clients should carefully analyze whether their price algorithms use similar inducement mechanisms to encourage other competitors to raise prices both on and off their own platforms.

Finally, given Amazon’s size and scope, clients should also evaluate whether they have been harmed by Amazon’s algorithms, or whether other companies using similar ones have caused harm.  Given the import of online superstores in our modern eCommerce economy, it is likely that algorithms did affect clients and change their behavior, potentially to the detriment of competition.  BCLP’s antitrust group can help clients understand and calculate those harms, as well as assess any risks.

*Note: BCLP Antitrust Partner David B. Schwartz was the FTC’s lead investigative attorney during the agency’s inquiry into Amazon’s eCommerce practices, however all of the information in this insight is based on publicly available information.