Amazon's $20B FTC Fight Tests Limits of AI-Powered Advertising

The Federal Trade Commission (FTC) is suing Amazon.com for allegedly manipulating its Sponsored Ads auctions, which could have inflated advertising costs by over $20 billion since 2019. The FTC and attorneys general from 22 states claim that Amazon used undisclosed 'soft reserve prices' to collect higher payments from advertisers.

Amazon disputes this, arguing that the FTC 'fundamentally misunderstands how advertisers operate.' It claims that its machine learning models improved ad relevance and conversion rates, resulting in average winning bids falling 50% between 2019 and 2025. The company also says that roughly 92% of sponsored ads shown to shoppers were not awarded to the highest bidder.

The case has implications beyond Amazon's advertising business, which generated $68 billion in revenue last year and is one of the company's fastest-growing, high-margin businesses. Regulators may challenge how algorithmic marketplaces operate, potentially subjecting AI-driven auction algorithms to disclosure and consumer protection standards similar to traditional pricing decisions.

Amazon insists that advertisers never pay more than their bid and claims the FTC relies on 'a handful of simplified communications' without evidence of consumer price increases. The outcome of this lawsuit could determine how courts treat AI-powered algorithms in digital marketplaces, with potential far-reaching implications for every platform where AI decides prices, rankings, and commercial outcomes.