Amazon Hits Perplexity With Legal Threat Over AI Bot

Amazon issued a stern cease-and-desist letter to AI startup Perplexity on Tuesday, demanding the company immediately stop its agentic shopping assistant, Comet, from browsing and making purchases on its e-commerce platform without properly identifying itself.

The Battle Over AI Agents: ‘Bullying’ vs. Transparency

The legal clash intensified after Perplexity published a blog post titled “Bullying is not innovation,” accusing the e-commerce giant of launching its first legal salvo against an AI firm. Perplexity argues that because its Comet assistant operates under the direct instruction of a human user, the bot should inherit the same access permissions as a human shopper. Consequently, the startup claims the AI does not need to declare itself as an automated agent while navigating Amazon’s digital storefront.

Amazon strongly disputes this interpretation, pointing to its terms of service which require automated systems to identify themselves. According to Amazon’s official response, transparency is standard industry practice. The retail giant noted that other third-party services—ranging from food delivery apps and digital personal shoppers to online travel agencies booking flights—routinely identify their bots to host platforms.

Ad Revenue and the Threat of the Rational Bot

While Amazon suggests that Perplexity could resume shopping activities simply by identifying its bot, doing so carries operational risks. Amazon operates its own proprietary shopping assistant, Rufus, and could choose to block Comet or competing third-party agents entirely once they are identified. Amazon’s statement emphasized that third-party applications making purchases on behalf of customers must operate openly and respect the decisions of service providers regarding whether to allow them on their platforms.

Perplexity contends that Amazon’s resistance is driven by financial motives, specifically the protection of its advertising and product placement revenue. Unlike human shoppers, who are highly susceptible to targeted ads, impulse buys, and sponsored recommendations, an AI agent programmed to purchase a specific item will execute the task with strict efficiency, bypassing Amazon’s highly profitable marketing loops.

var playerInstance_jwplayer_6a6938fea8dfd = jwplayer( “jwplayer_6a6938fea8dfd” );
playerInstance_jwplayer_6a6938fea8dfd.setup({
playlist: “https://cdn.jwplayer.com/v2/media/nsQAyeWN”,
});

A Recurring Pattern of AI Hiding Tactics

This is not the first time Perplexity has faced scrutiny over its data collection methods. Earlier this year, cybersecurity firm Cloudflare published research revealing that Perplexity was actively scraping websites while ignoring direct “do not crawl” requests. During that controversy, defenders argued that the AI was merely fetching real-time data to answer specific user queries, mimicking standard web browser behavior. However, evidence showed the AI startup utilized masking techniques to hide its identity when accessing restricted domains.

The Future of Agentic E-Commerce

The dispute highlights a growing challenge for the future of the internet. As consumers increasingly delegate tasks like shopping, travel bookings, and reservations to autonomous AI agents, platforms must decide how to interact with these non-human users. By taking a hardline stance, Amazon is establishing a major industry precedent: AI agents must identify themselves, leaving the ultimate control over access in the hands of the platform owners.

Following the public debate, Amazon has also published the complete cease-and-desist letter sent to the AI startup, detailing its legal arguments.