A U.S. federal appeals court has overturned a preliminary injunction that blocked Perplexity from using its AI agent Comet for automated purchases on Amazon. This is a landmark decision that, for the first time at the appellate level, clarifies the legal status of autonomous algorithms' actions in the web environment.
The court's key argument is that responsibility for Comet's actions lies with the user who launches the agent, not with Perplexity itself. The judges found that the AI assistant acts as a tool, not an independent entity performing legally significant operations. Thus, Amazon failed to prove that Perplexity as a platform violates the service's terms of use.
This ruling creates an important precedent for the entire industry. Essentially, the court recognized that an AI agent's interaction with websites is not automatically equated with the developer's actions. For Amazon, this is a serious blow to its attempts to restrict automated systems' access to its marketplaces — especially in light of the growing popularity of AI assistants for online shopping.
However, it is worth emphasizing: the court only overturned a preliminary measure, not issued a final ruling on the merits of the dispute. The litigation between Amazon and Perplexity continues, and the final verdict may differ. Nevertheless, it is already clear that regulatory and judicial practice is beginning to adapt to a new reality where AI agents are becoming full-fledged participants in the digital economy.
My comment: This is only the first step in a long legal battle. If the final ruling cements this approach, we will see an explosive growth in the number of AI agents acting on behalf of users in commercial services — but this will also raise new questions about security and fraud. Investors should closely monitor the development of this case: it could reshape the rules of the game for the entire automated assistant market.