The 9th U.S. Circuit Court of Appeals in San Francisco dealt Amazon a significant legal setback on Tuesday by overturning a lower court's ban on Perplexity's artificial intelligence shopping agents from operating on the e-commerce giant's platform. The decision marks a watershed moment in US technology law, as it represents the first major federal appeals court ruling addressing whether autonomous AI systems can legally access commercial websites and conduct transactions on behalf of users without explicit platform authorization.
Amazon had initiated legal proceedings against the AI startup in November, alleging that Perplexity had been systematically accessing private customer accounts through its Comet browser and integrated AI agents without permission. According to Amazon's complaint, the system allowed users to authenticate their existing shopping credentials and authorize the AI to browse their accounts, view purchase histories, and even execute transactions independently. The company contended that these activities created security vulnerabilities and violated the Computer Fraud and Abuse Act, a sweeping 1986 statute designed to protect computer systems from unauthorized access. Amazon emphasized that it had repeatedly requested Perplexity cease the practice, only to have its demands ignored.
Perplexity mounted a spirited defence, characterizing Amazon's lawsuit as a transparently protectionist move designed to prevent consumers from leveraging competing AI tools. The startup's response argued that the case lacked legal merit and represented nothing more than an attempt to lock users into Amazon's preferred ecosystem. In a pointed observation about digital advertising, Perplexity suggested that Amazon's true concern stemmed from the fact that its AI agents "don't have eyeballs to see the pervasive advertising Amazon bombards its users with," implying the company feared revenue loss from reduced ad exposure rather than genuine security concerns.
The initial outcome appeared to vindicate Amazon's position when a federal judge in California issued a temporary restraining order in March, finding that the e-commerce company had presented "strong evidence" supporting its claim that Perplexity's activities constituted a violation of federal computer fraud statutes. This preliminary injunction effectively halted Perplexity's shopping functionality on Amazon's platform, creating a tangible consequence for the AI startup's business operations.
However, the appeals court's reversal fundamentally reframed the legal question at stake. Rather than viewing Perplexity as the entity accessing Amazon's computers, the three-judge panel determined that responsibility lay with individual users who had explicitly authorized the AI agent to act on their behalf. This distinction proved dispositive: the Computer Fraud and Abuse Act, the court reasoned, targets unauthorized access by the entity itself, not access undertaken by authorized users employing technology intermediaries. By characterizing Perplexity's role as analogous to that of a browser or software tool rather than as an independent actor, the court opened a considerably wider door for agentic AI systems.
The implications of this ruling extend well beyond Amazon and Perplexity's commercial dispute. Agentic AI represents an emerging category of artificial intelligence designed to plan, reason and execute complex tasks across the internet with minimal human intervention. These systems can navigate multiple websites, authenticate with various platforms, gather information, make purchasing decisions, and facilitate transactions—all in service of user objectives. The decision effectively legitimizes this model as a matter of law, at least within the 9th Circuit's jurisdiction, which covers the western United States and holds substantial influence over technology policy.
Amazon responded to the defeat with measured resistance, stating through a company spokesperson that it "respectfully disagreed" with the appeals court's analysis and remained "confident in our case" while evaluating potential next steps. These comments suggest the possibility of further appeals, potentially to the Supreme Court, though such a petition would face long odds given the appellate court's thorough analysis. The company faces a difficult strategic choice: pursuing more litigation at considerable expense with uncertain prospects, or accepting a market reality in which AI agents can access its platform provided users have authorized such access.
Perplexity's leadership celebrated the decision as validation of their broader philosophy regarding AI accessibility and user choice. Jesse Dwyer, a company spokesperson, emphasized that Perplexity would "continue to fight for the right of internet users to choose whatever AI they want," framing the dispute in consumer-protective terms rather than as a narrow commercial squabble between two technology companies. This rhetorical positioning reflects how agentic AI companies view their role—as enablers of user autonomy rather than as unauthorized system intruders.
The decision carries profound implications for Southeast Asian digital ecosystems and Malaysian consumers specifically. As agentic AI tools proliferate globally, regulators and platforms across the region will face mounting pressure to clarify their own positions on whether and how such agents can operate within their boundaries. The legal framework emerging from US courts influences how technology companies approach compliance worldwide, meaning that this ruling will likely shape platform policies and feature availability in Malaysia and neighbouring countries.
The ruling also highlights a fundamental tension in contemporary technology governance between platform control and user empowerment. Amazon's position prioritized its ability to govern access to its proprietary systems and customer data, emphasizing security and contractual rights. The appeals court instead emphasised the legitimate interests of users in deploying tools—including AI agents—to accomplish their objectives on platforms where they already have accounts. This user-centric approach may resonate differently across various jurisdictions and regulatory philosophies, particularly in regions where data protection and consumer rights remain areas of active legislative development.
For the broader agentic AI industry, the decision represents a crucial legal validation. Numerous startups are developing autonomous agents that can browse the web, manage email accounts, coordinate across multiple platforms, and execute transactions. Prior to this ruling, the legal environment surrounding such capabilities remained deeply uncertain, potentially chilling innovation. The appeals court's reasoning provides these companies with considerably stronger legal footing to build and deploy sophisticated agents, assuming they can demonstrate that users have provided informed consent.
Looking ahead, observers expect this case to generate substantial follow-up litigation as major platforms attempt to reassert control over how third parties access their systems. The stakes extend beyond shopping functionality to encompass email management, financial services access, social media engagement, and countless other domains where agentic AI could operate. How courts ultimately balance platform property rights against user agency in the context of autonomous AI will shape the digital landscape for years to come.
