The San Francisco-based 9th U.S. Circuit Court of Appeals has delivered a significant victory to artificial intelligence startup Perplexity, striking down a temporary court ban that had prevented the company from deploying its autonomous shopping agents on Amazon's platform. The decision, handed down on Tuesday, represents the first major federal appeals ruling to directly address whether AI agents operating on behalf of users can legally access online platforms without explicit permission, establishing precedent in an emerging and contentious area of technology law.
Amazon initiated legal proceedings against Perplexity in November, alleging that the startup's Comet browser and associated AI agent were illicitly accessing private customer accounts without authorisation. According to the lawsuit, Perplexity's systems could log into users' shopping profiles and execute purchases independently, creating security vulnerabilities while circumventing Amazon's own technological infrastructure. Amazon claimed it had repeatedly requested that Perplexity cease these activities, but the company persisted in developing and deploying the capability, prompting the tech giant to seek judicial intervention.
In March, a federal judge in California initially sided with Amazon, issuing a temporary restraining order after finding what the judge described as "strong evidence" that Perplexity had violated the Computer Fraud and Abuse Act, a cornerstone piece of American cybersecurity legislation. However, the appeals court's reversal on Tuesday fundamentally reframed the legal analysis, adopting Perplexity's interpretation of how the relevant statute applies to AI-mediated access. The 9th Circuit concluded that under the law's language, it is the users themselves—operating through Perplexity's AI agents as their chosen tools—who access Amazon's servers, rather than Perplexity acting independently as the accessing party.
This distinction carries profound implications for the rapidly evolving sector of agentic AI systems, which represent a new category of autonomous software capable of planning, reasoning, and executing complex tasks with minimal human intervention. These tools can navigate websites, make purchasing decisions, initiate transactions, and gather information across the internet without requiring users to manually perform each action themselves. The legal question of whether such systems operate as independent agents or merely as extensions of user intent has become central to understanding how AI will integrate into everyday digital commerce and information gathering.
Perplexity's representatives expressed confidence in the outcome, with company spokesperson Jesse Dwyer framing the ruling as a vindication of user rights. The company characterised the dispute as fundamentally about consumer choice, arguing that internet users should retain the freedom to employ whatever AI tools they prefer for their online activities. Perplexity's earlier response to Amazon's allegations contended that the lawsuit represented a "bald attempt" to monopolise its own platform by preventing AI agents from accessing it, particularly because such agents lack the capacity to process the advertising that Amazon embeds throughout its shopping ecosystem.
Amazon responded to the appellate decision with measured language indicating continued confidence in its legal position. A company spokesperson stated the organisation "respectfully disagreed" with the appeals court's reasoning and announced it was evaluating subsequent legal options, leaving open the possibility of further appeals or alternative legal strategies. Amazon's resistance to AI agents accessing its platform reflects broader concerns among major technology companies about maintaining control over how their services are accessed and monetised, particularly as AI capabilities expand to automate user interactions previously requiring human attention.
The underlying dispute touches on fundamental tensions in digital commerce between platform owners seeking to preserve their business models and emerging AI companies arguing for open access to publicly available information and user-authorised transactions. Amazon's advertising-dependent revenue model depends significantly on users viewing product recommendations and sponsored listings, something AI agents programmed for efficiency might bypass entirely. Perplexity's assertion that AI agents "don't have eyeballs to see the pervasive advertising Amazon bombards its users with" highlights how autonomous systems could substantially disrupt traditional digital advertising architecture.
For Southeast Asian technology observers and policymakers, this ruling carries particular relevance as the region grapples with regulating AI development while remaining competitive in the global technology sector. Malaysian and regional entrepreneurs building AI applications face uncertainty about what autonomous capabilities will be legally permissible when deployed internationally, particularly when interacting with major platforms headquartered in the United States. The decision suggests American courts may be more permissive of AI agent access than technology companies would prefer, though the question remains unsettled at the highest judicial levels.
The 9th Circuit's reasoning creates a precedent that could embolden other AI developers to build autonomous agents capable of accessing major platforms under the legal theory that user intent, rather than the technology itself, constitutes the actionable access. This could accelerate development of agentic AI tools across shopping, research, and other domains where autonomous access currently faces legal uncertainty. However, the decision is not final, and Amazon's indication that it is evaluating next steps suggests this dispute may continue through additional appeals or legislative action, potentially reaching the Supreme Court if the company pursues further review.
The broader implications extend beyond any single company's dispute. As agentic AI systems become more sophisticated and widespread, questions about their legal status—whether they function as independent actors or user-delegated tools—will shape the entire emerging ecosystem of autonomous AI applications. The 9th Circuit's decision to treat user agency as controlling in AI interactions may influence how other courts approach similar disputes, potentially establishing permissive precedent for AI agent deployment across multiple platforms and industries.
