Do Son 
Perplexity, a pioneering artificial intelligence search engine and agentic service startup, has secured a monumental reversal in its legal confrontation with e-commerce titan Amazon regarding automated AI shopping capabilities.
The United States Court of Appeals for the Ninth Circuit has officially dissolved a previously issued injunction. This ruling permits the AI shopping assistant embedded within Perplexity’s Comet browser to access Amazon’s digital storefront once more. Consequently, the Perplexity AI shopping bot can resume conducting price comparisons and executing purchases on behalf of users across the Amazon platform.
The Judicial Core: User Agency Versus Corporate Intrusion
Initially, Amazon contended that Perplexity’s Comet browser deployed AI bots to autonomously scrape proprietary website data and execute unauthorized operations. Amazon alleged this behavior blatantly violated the United States Computer Fraud and Abuse Act (CFAA).
However, the appellate court’s latest ruling explicitly dismantled Amazon’s accusations.
To establish a CFAA violation, a plaintiff must demonstrate that the defendant intentionally accessed a protected computer without authorization, resulting in tangible damages. Because the Comet browser’s AI shopping function requires an explicit user directive to navigate Amazon’s website, the true entity accessing the servers is the individual user, rather than the corporate entity of Perplexity.
The court determined that issuing an injunction against behavior that likely does not violate the CFAA contradicts the public interest. Furthermore, any potential detriment Amazon might suffer from Comet’s operation remains insufficient to justify the imposition of a legal injunction.
Addressing this judicial reversal, Amazon articulated in a formal statement: “We respect, but fundamentally disagree with today’s ruling regarding the preliminary injunction. We retain absolute confidence in our litigation posture and are actively evaluating our subsequent maneuvers.”
The Evolution of Conflict: From Clandestine Agreements to Courtroom Battles
This escalating conflict over agentic AI shopping traces its origins to 2024. During that period, the two enterprises forged a temporary pact to suspend AI proxy shopping. Nevertheless, Amazon later accused Perplexity of brazenly violating this agreement. They alleged that Perplexity deliberately disguised Comet’s AI bots as standard Chrome browsers to circumvent Amazon’s defensive blockades.
- November 2025: Amazon formally issued a cease-and-desist letter to Perplexity, demanding the immediate cessation of AI browser access to its digital storefront.
- March 2026: Amazon initiated a lawsuit in a San Francisco federal court and subsequently secured a preliminary injunction, forcibly blocking the Comet browser.
- August 2026: The United States Court of Appeals for the Ninth Circuit unequivocally revoked the aforementioned injunction.
As noted in reports detailing how Perplexity overturns the Amazon ban on AI shopping bots on appeal, although the appellate court dissolved the injunction, the overarching litigation remains far from concluded. Amazon retains the prerogative to petition for an en banc rehearing or to seek a comprehensive review by the United States Supreme Court. Even if Amazon eschews these appellate avenues, the primary lawsuit will proceed unabated within the San Francisco federal court.
Concluding Perspectives
This specific ruling establishes an exceptionally critical benchmark for the rapidly burgeoning agentic AI industry.
As AI agents like Gemini Spark and Perplexity Comet increasingly acquire the capacity to autonomously navigate the web, complete forms, compare prices, and execute orders on behalf of users, traditional e-commerce platforms confront unprecedented challenges. Their established defenses regarding website traffic control and user data surveillance face profound disruption. Naturally, platform operators like Amazon strive to preserve their closed ecosystems. They aim to prevent third-party AI from siphoning users and proprietary data without generating direct advertising revenue.
However, the court legally characterized the actions of AI agents as an extension of the user. The ruling emphasized that as long as a user provides explicit authorization and directives, an AI agent accessing a public website does not constitute an illegal intrusion. This landmark decision paves the way for AI browsers and agentic tools, effectively preventing traditional technological titans from abusing CFAA injunctions to suppress nascent innovation.
Nevertheless, as the primary litigation advances, the legal identity and jurisdictional boundaries of AI agents within the digital realm will undoubtedly remain the most fiercely contested battleground in technology law for years to come.
Support Our Threat Intelligence
If you find our CVE report and cybersecurity news helpful, consider supporting our work.
Buy Me a Coffee
PayPal
