A different kind of case against Amazon
On August 31, 2026, the Federal Trade Commission and attorneys general from 22 states sued Amazon in the U.S. District Court for the Western District of Washington, alleging that since 2019 the company ran a covert surcharge scheme inside its advertising auctions. The complaint, authorized by a 2-0 Commission vote, says Amazon told advertisers it was running a textbook "second-price" auction — where the winning bidder pays one cent more than the second-highest bid — while secretly layering in an undisclosed "soft reserve price" and what internal documents reportedly call an "invented auction participant." The upshot, per the FTC: Sponsored Products advertisers were charged their own full winning bid roughly 80% of the time, turning a marketed second-price mechanism into a de facto first-price one, without telling more than a million brands and sellers who bid on that promise (FTC press release; TechCrunch).
This is worth pausing on, because it is a narrower and legally cleaner claim than the FTC's other pending fight with Amazon. In September 2023, the agency and 17 states sued Amazon under Section 5 of the FTC Act and Section 2 of the Sherman Act, alleging the company illegally maintains monopoly power over online retail — a sprawling structural case now headed toward trial (FTC). The ad-auction suit doesn't ask a court to referee what "fair" market structure looks like. It alleges Amazon said X and did Y — a disclosure and deception claim, closer in kind to the Prime "dark patterns" case the company settled for $2.5 billion in September 2025 ($1 billion in penalties, $1.5 billion in consumer refunds) than to a Sherman Act monopolization theory (Time).
Steelmanning the case
The strongest version of the FTC's argument doesn't require believing Amazon is a monopolist, or that ad auctions should be regulated. It requires believing that if a company tells 500,000+ small and medium-sized businesses "you'll pay one cent more than the next bidder," and then quietly injects a phantom minimum bid that pushes most winners up to their own ceiling, that is a material misrepresentation advertisers priced their bidding strategy around. Auction design is not incidental to price discovery — it's the mechanism by which price discovery happens. A seller who believes they're in a second-price auction bids more aggressively than one who knows the house can effectively see their number and charge it back to them. If the FTC's allegations are accurate, Amazon's sellers — many of them small merchants with thin margins already squeezed by referral and fulfillment fees — were bidding blind against a rule Amazon had unilaterally rewritten and did not disclose. That is a serious allegation, and if proven, straightforward fraud on the inducement, not an overreach of antitrust theory into ordinary commercial behavior.
Why the frame still matters
But precisely because this is a disclosure case and not a structural one, the remedy should track the harm. The complaint reportedly seeks civil penalties, restitution, and injunctive relief — not a breakup, divestiture, or line-of-business separation. That is the correct scope. Advertising auctions are opaque by design across the entire industry — Google, Meta, and every programmatic ad exchange run pricing logic that advertisers cannot fully audit, and "soft floors" and reserve pricing are common, disclosed tools in auction theory generally. The FTC's theory here isn't that Amazon used a reserve price — reserve prices are legitimate and widely used to prevent auctions from clearing below a seller's minimum acceptable value. The theory is that Amazon told advertisers it wasn't using one, or was using a materially different mechanism than the one actually running. That distinction is what should anchor both the litigation and the commentary around it: courts and regulators should be exacting about proven misrepresentation, and correspondingly restrained about extrapolating from one company's alleged deception to a broader indictment of algorithmic pricing or reserve-price auctions as such.
That restraint cuts against Amazon too. "Price variations naturally occur across billions of bids," the company said in response, arguing the FTC "fundamentally misunderstands how advertisers operate." That may be true of ordinary auction noise. It is a much harder sell for a mechanism specifically internally labeled a reserve price and an invented bidder — if those documents say what the FTC claims, "noise" is not the right word for a deliberately engineered floor. Amazon generated more than $68 billion in advertising revenue in the year preceding this suit; a company at that scale, running the marketplace and the auction simultaneously, owes advertisers auction mechanics that match its own marketing claims. Discovery, not press statements, will settle which account is right.
What proportionate regulation looks like here
The useful signal for a pro-innovation, pro-competition read of this case is process, not outcome-shopping. A disclosure claim, tested at trial or settled on the facts, is exactly how consumer-protection law is supposed to work — it punishes concealment without requiring a court to redesign a business model. That's a healthier template than the FTC reaching for antitrust remedies every time a large platform's internal practices diverge from its public claims. If Amazon misrepresented its auction to a million businesses, restitution and a court-enforced disclosure regime are proportionate; a monopolization theory would not be necessary to get there, and this case doesn't require one. The real test of proportionality will come at the remedy stage — whether the FTC asks for what the alleged deception actually cost advertisers, or reaches for something bigger because the target is Amazon.