TLDRs;
- Amazon sent a legal warning to Perplexity AI over its Comet browser’s automated shopping feature.
- Perplexity accuses Amazon of anticompetitive behavior and limiting innovation in agent-driven e-commerce tools.
- The dispute highlights unclear global rules around AI agents performing online transactions autonomously.
- Experts say API-based compliance could offer a middle ground for AI-assisted shopping in the future.
Amazon has formally warned Perplexity AI to disable automated purchasing capabilities in its Comet browser, claiming that the feature violates Amazon’s Conditions of Use.
The e-commerce titan argued that Comet’s ability to log in, browse, and complete purchases on behalf of users without direct consent from Amazon could lead to a “degraded shopping and customer service experience.”
Perplexity AI, a U.S.-based startup known for building advanced AI assistants, launched Comet earlier this year as a next-generation browser designed to act as an “autonomous agent.” The browser enables users to automate online tasks, including purchasing items, using AI commands. Amazon, however, maintains that such agentic activity crosses the line into unauthorized automation, prohibited under its existing platform rules.
According to Amazon’s statement, “Third-party applications that place orders on behalf of customers must operate transparently and comply with the terms of service of all platforms they access.”
Despite repeated requests, Amazon claims that as of November 5, Perplexity had not removed the company’s platform from the Comet experience.
Perplexity Pushes Back, Claims Market Suppression
In response, Perplexity AI accused Amazon of using its market dominance to suppress emerging competition in AI-driven commerce. The startup argues that Comet merely streamlines user intent rather than violating platform integrity.
“Comet does not scrape or mine data, it simply assists users in executing their own shopping preferences,” Perplexity said in a statement.
The company suggested Amazon’s actions reflect an industry-wide fear of AI autonomy disrupting traditional retail intermediaries.
The standoff revives a broader debate: Should AI agents be treated as legitimate digital proxies for users, or are they effectively bots violating website rules? With few legal precedents, experts suggest that the issue may ultimately need to be settled in court.
Legal Gray Zone for AI Shopping Agents
The controversy underscores how the law is struggling to catch up with technological innovation.
Under Amazon’s Conditions of Use, customers are prohibited from using “robots, data mining tools, or similar mechanisms” to interact with the site. Comet, however, exists in a gray area, its automation features mimic human behavior while still requiring user permission.
Earlier warnings between Amazon and Perplexity date back to late 2024, when both sides agreed to pause “agentic shopping” until clearer boundaries could be established. But when Comet’s latest version relaunched, Perplexity reportedly reactivated the feature, masking its operations to appear like a standard Chrome browser.
Toward a Compliant Future for AI Commerce
Industry watchers say the dispute may accelerate calls for standardized APIs for AI-driven commerce. Platforms like Shopify already offer a Storefront API, allowing developers to create custom shopping agents that operate transparently with merchant consent.
Such API-based approaches could become the foundation for compliant “agent shopping,” allowing startups to innovate without breaching service terms. Investors are also being urged to back companies that work cooperatively with retailers rather than testing legal boundaries.
Amazon has hinted that it’s open to collaboration,provided third parties use approved checkout mechanisms that maintain customer safety and transparency. As AI continues to blur the lines between user intent and autonomous action, the Perplexity-Amazon conflict may set a critical precedent for the next phase of online retail.


