A federal judge has granted a temporary court order blocking the artificial-intelligence startup Perplexity from deploying its automated shopping bots on a major online retail platform after the company argued the bots were accessing customer accounts without authorization and bypassing platform safeguards. The ruling stems from a lawsuit filed in late 2025 alleging that the AI startup’s “Comet” browser and its agentic shopping tool secretly automated purchases, scraped product data, and disguised bot activity as normal human browsing. In issuing the injunction, the court found strong evidence that the software entered password-protected areas of the retailer’s systems without the platform’s consent, even if users themselves authorized the AI to act on their behalf. The temporary block requires the AI firm to halt use of its shopping agents on the retailer’s website and destroy any data collected from the platform while litigation continues. The dispute underscores a rapidly intensifying battle over who controls access to online marketplaces in the emerging age of autonomous AI assistants that can browse, compare prices, and complete purchases on behalf of consumers.
Sources
https://www.theepochtimes.com/business/amazon-wins-court-order-to-block-perplexity-ai-shopping-bots-5996790
https://www.reuters.com/legal/litigation/amazon-wins-order-blocking-access-perplexitys-ai-shopping-agent-2026-03-10/
https://www.theverge.com/ai-artificial-intelligence/892401/amazon-perplexity-ai-shopping-agent-court-order
Key Takeaways
- A federal judge issued a preliminary injunction blocking the AI startup’s automated shopping agents from accessing a major e-commerce platform after evidence suggested the bots entered password-protected accounts without authorization.
- The court ruling represents one of the first major legal tests over whether AI assistants can autonomously interact with third-party websites and complete transactions without explicit approval from the platform owner.
- The dispute highlights a growing power struggle between large technology platforms and emerging AI companies over control of user data, marketplace access, and the future of automated online commerce.
In-Depth
The legal clash between a dominant online retailer and the AI startup Perplexity marks one of the earliest courtroom showdowns over “agentic commerce,” the concept of artificial-intelligence tools acting as digital assistants that can independently browse the internet and complete tasks such as shopping. At the center of the dispute is Perplexity’s Comet browser, which integrates an AI agent designed to search for products, compare prices, and place orders automatically on behalf of users. While the company framed the feature as a convenience that empowers consumers, the retail platform argued it crossed a legal line by accessing its systems in ways that violated both its terms of service and federal computer fraud laws.
The retailer filed suit in November 2025, alleging that the AI software covertly entered customer accounts and masked automated activity so it would appear indistinguishable from normal human browsing. According to court filings, the bots could navigate password-protected areas of the platform and make purchases using a user’s credentials. The company contended that this behavior not only threatened customer data security but also undermined the carefully structured experience it provides shoppers. By automating the process, the bots could bypass advertising placements, recommendation systems, and other mechanisms the retailer uses to manage its marketplace.
In granting a preliminary injunction, the judge determined that the retailer had presented strong evidence that the AI system accessed its platform without authorization from the company itself, even if individual users permitted the AI to act on their behalf. That distinction proved pivotal. The ruling essentially states that control over a digital platform ultimately rests with the platform owner, not third-party developers who attempt to automate interactions with it.
For the AI industry, the decision could set an early precedent in the race to build autonomous web agents capable of handling everyday tasks. Companies across Silicon Valley are investing heavily in tools that allow AI assistants to schedule travel, draft messages, and conduct purchases without human intervention. Yet the ruling signals that traditional tech giants are prepared to push back aggressively if those agents interfere with their business models or gain access to proprietary data.
Perplexity has indicated it plans to challenge the order, arguing that consumers should be free to use whatever software tools they choose to interact with online services. The company has portrayed the lawsuit as an attempt by an entrenched technology power to choke off innovation and protect its commercial ecosystem. Critics of the platform’s position also argue that if users willingly provide credentials to an AI assistant, it should be treated no differently than any other tool that automates web activity.
Still, the broader implications are difficult to ignore. If autonomous AI agents become widespread, they could fundamentally alter how digital marketplaces function. Shopping could shift away from human browsing toward algorithm-driven decision-making, where software quickly compares thousands of options and executes purchases instantly. Such a future would challenge existing revenue models built around advertising, product placement, and curated user experiences.
For now, the court’s decision draws a clear line: AI companies cannot simply deploy autonomous agents inside another company’s digital ecosystem without permission. As lawsuits multiply and regulators begin paying closer attention, the emerging rules governing AI-driven automation may ultimately determine who controls the next generation of the internet—those who build the platforms, or those who build the machines that navigate them.