Perplexity AI agent wins appeal over Amazon CFAA claims
A U.S. appeals court dealt an early legal win to Perplexity in its dispute with Amazon over the startup’s AI shopping agent on Amazon’s site. The Ninth Circuit Court of Appeals ruled Tuesday that Amazon is unlikely to prove Perplexity violated the Computer Fraud and Abuse Act (CFAA).
Key points:
- The court said Amazon failed to show the required “without authorization” element under CFAA. It found that Amazon’s users, not Perplexity, accessed Amazon through Perplexity’s agentic AI browser “Comet”.
- Comet can browse websites, log into user accounts, and complete purchases on a user’s behalf.
- The legal fight began after Amazon sent Perplexity a cease-and-desist letter in November, then sued.
- Amazon argued Comet disguised automated traffic as a standard browser and accessed customer accounts in violation of its terms.
- Perplexity said Comet only acted with users’ permission, using credentials stored on the users’ own devices, and accused Amazon of trying to restrict consumer choice.
- In March, a federal judge granted Amazon a preliminary injunction, temporarily blocking Comet from shopping on Amazon while the case proceeded.
While the underlying lawsuit remains ongoing, the appellate ruling signals how courts may treat agentic AI systems that act on users’ behalf in online commerce.
Neutral
This is primarily a U.S. legal/compliance ruling for agentic AI in online commerce, not a crypto protocol change or token-specific decision. The Ninth Circuit’s narrowing of Amazon’s CFAA theory reduces near-term uncertainty for Perplexity’s agent shopping functionality, which is mildly sentiment-positive for the “AI agents” theme. However, it has no direct linkage to major crypto assets or to blockchain infrastructure flows.
Historically, market impact from tech-sector court rulings has usually been limited unless it affects a widely held public platform, changes monetization at scale, or triggers a broader regulatory shift. Here, the ruling is an early appellate decision while the underlying case continues, so traders are likely to treat it as background noise rather than a catalyst.
Short term: neutral, with any effect more likely to be seen in AI-related equities/venture sentiment than in BTC/ETH.
Long term: neutral-to-slightly constructive for the legal feasibility of agentic AI, which can indirectly support Web2/Web3 integrations later—but not enough for a crypto market re-pricing on its own.