The FTC’s third major federal action against Amazon — filed jointly with a bipartisan group of 22 state attorneys general in U.S. District Court in Washington — targets the company’s advertising business, which earned almost $69 billion in 2025 and grew 26 percent in the most recent quarter. The complaint alleges Amazon “secretly and systemically overcharging” advertisers and that “in executing its scheme to misrepresent and rig its auctions against its customers, Amazon has deceived its customers and deprived them of the benefits of competition and fair and transparent dealing.” Amazon disputes the allegations, characterizing the FTC’s case as built on cherry-picked findings and citing data showing the average cost-per-click on its ads held flat when adjusted for inflation.

The government’s complaint and Amazon’s public response do not engage the same claim. The complaint targets the auction process — that Amazon raised the prices brands and sellers pay by manipulating bidding mechanics. Amazon’s defense addresses outcomes: cost-per-click stability in real terms, ad-placement quality improving over time, and no documented harm to shoppers, who do not pay for ads. A flat average cost-per-click does not preclude auction-level manipulation; it could be consistent with manipulation that redistributes costs across advertisers or inflates particular placements while leaving the average unchanged. Improved placement quality is consistent with either legitimate auction optimization or with manipulation that makes higher bids appear necessary. The “no harm to shoppers” point redirects from the complaint’s actual target — brands, sellers, and advertisers, not end consumers. Amazon’s additional framing that “the auction format Amazon uses to sell ads is fairly common in the industry” introduces a third claim — that the practice, whatever it is, is not Amazon-specific — which the government could in turn deploy to argue industry-wide harm if substantiated.

Three competing explanations are consistent with the public record. The first is that Amazon engaged in undisclosed manipulation as the complaint describes. The second is that Amazon’s ad-auction changes were disclosed optimizations. The third is that the specific mechanisms Amazon uses are common across major ad platforms, so any finding would be industry-wide rather than Amazon-specific. Assessing diagnosticity: the government’s allegation that the conduct was “secretly” executed cuts against H2 and toward H1 if substantiated; Amazon’s claim of a flat real-terms cost-per-click cuts against a strong version of H1 in which manipulation produced price inflation at the average level but is consistent with weaker versions in which manipulation redistributed costs among advertisers; the “common in industry” framing cuts toward H3; and the 26 percent quarterly growth in ad revenue is consistent with all three and therefore low-diagnosticity. Resolving among the three would require access to internal Amazon documents on auction design, advertiser communications, and the specific mechanisms the complaint identifies — material not present in public reporting. What would falsify each: H1 fails if internal records show disclosed optimization; H2 fails if records show undisclosed manipulation; H3 fails if Amazon’s mechanisms differ materially from peers’. The conclusion that the public record does not yet discriminate among the three is a coverage gap, not an inference in favor of any party.

The consequences extend across several time horizons. Immediate: the filing adds a third federal action to the FTC and 17-state monopolization case scheduled for trial early next year and the $2.5 billion Prime settlement reached just over a year ago, altering the settlement-versus-litigation calculus across the active proceedings. Short-term: the 1.2 million brands and sellers identified in the complaint represent a potential follow-on class-action base; private litigation frequently follows government actions of comparable scope. Medium-term: the case joins an emerging regulatory focus on ad-auction transparency across the technology sector, with the “common in industry” framing potentially inviting coordinated examination of peer platforms. Longer-term: an adverse finding in the ad case could compound with remedies sought in the monopolization case — which the FTC and 17 states have brought on grounds that Amazon “suffocates competitors and raises costs for both sellers and shoppers” — potentially reaching the structural level of how Amazon integrates advertising and marketplace operations. A reinforcing branch runs through Amazon’s earlier resolutions (the $2.5 billion Prime settlement and the 2023 $30 million Alexa and Ring privacy settlement), which establish a documented pattern of regulatory engagement that makes each subsequent action more procedurally and politically consequential. A counteracting branch runs through Amazon’s “industry-common” framing, which if upheld could narrow the case to its specific mechanisms rather than producing broader precedent.

The three-action pattern indicates an enforcement architecture rather than isolated cases. Whether that architecture produces structural remedies or a series of monetary resolutions depends on the specific findings in each proceeding and on how the cases interact procedurally.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Argument Audit
A full structural audit of an argument’s premises, inferences, and load-bearing assumptions.
Analysis of Competing Hypotheses
Scores rival explanations by how well each fits the evidence, weighting the diagnostic items (Heuer).
Consequences & Sequels
Plays a decision forward to its first- and second-order consequences.
Confirmation Bias
Seeking and overweighting the evidence that confirms what one already believes.
Winner’s Curse
In a contested auction, winning often means having overpaid.