A U.S. federal appeals court has made a landmark decision, overturning a preliminary injunction that blocked the integration of Perplexity's AI agent Comet with Amazon's shopping platform. This is the first such precedent at the appellate level, affecting the legal status of autonomous AI assistants when interacting with web services.

The core of the dispute was whether Comet's use for automated purchases violates Amazon's terms of service. However, the panel of judges reached a different conclusion: the key argument was that the actions are performed by the end user, delegating authority to their digital assistant, rather than by Perplexity itself as the service operator. Thus, there are insufficient grounds for such a harsh preventive measure as an injunction.

Legal vacuum and the future of AI agents

This decision opens a new chapter in legal practice concerning AI agents. It effectively recognizes that automated tools act on behalf of the user, not as independent entities violating platform rules. This creates an important precedent for dozens of similar disputes that will inevitably arise as such technologies spread in e-commerce and other areas.

It is important to emphasize that the litigation between Amazon and Perplexity is far from over—the lifting of the injunction does not mean a final victory for the defendant. The final ruling on the merits of the claim is still ahead, and it could significantly adjust the current legal landscape.

My analysis: This decision is a signal to the market that courts are not yet ready to fully block innovation in the field of AI agents, but it also creates a risk for platforms trying to control access to their data. In the long term, we will likely see either stricter user agreements or new legislative initiatives regulating this segment. Investors and developers should closely monitor the case's progress—it could define the standards for interaction between AI and web platforms for years to come.