If you let an AI do your shopping, can a court actually rule on whether that's allowed?

This explores whether courts can rule on whether handing your shopping to an AI agent is legally allowed. The corpus suggests they can, but only in narrow pieces. The clearest example is a single ruling that sidestepped the broad question. When Amazon tried to block Perplexity's shopping agent under the federal anti-hacking law, the Ninth Circuit vacated the injunction. Its reasoning was technical: the user's own browser talks to Amazon, and Perplexity's servers only analyze screenshots, so legally the user is the one accessing Amazon's computers Who accesses a website when an AI agent browses it?. In effect, the court treated the agent as an extension of the person. That's a meaningful win for delegation, but it answers "who touched the server?", not "is delegation permissible?"

The ruling also leaves a lot open. Retailers don't need a court to stop agents. When Amazon blocked Meta's Muse agent, it did so through its terms of service, because there is no standard way for an agent to identify itself or prove it has the shopper's permission What identity standards do shopping agents actually need?. Until those identity and authorization protocols exist, platform rules decide access more often than judges do. That's why Senator Warner's draft AI AGENT Act tries to settle the issue through legislation. It would require large platforms to admit agents, create an FTC registry of trusted agents, and have NIST set technical standards Could federal AI agent registries solve platform access unfairness?.

The deeper legal question may be less about whether you can delegate and more about what the agent owes you once you do. Owens and Law argue that AI shopping agents should carry the same fiduciary duties (legal duties of loyalty and care) that apply to human brokers. They also warn that the window to impose these duties is closing as Google, OpenAI, and Robinhood scale up agent features Should AI shopping agents face fiduciary duties like human brokers?. This matters because loyalty is already under strain. Azhar points out that an agent paid through merchant referral fees can't be loyal to both the merchant and the shopper Can an AI agent serve both merchant and user interests fairly?. In an experiment, an LLM agent told that a booking platform was its employer penalized sponsored listings 21 points less than one told it worked for the traveler Does naming the platform change how AI agents judge sponsored listings?. Whose agent it is changes how it behaves.

Europe shows a different gap: the laws exist, but they assume a human is shopping. EU consumer protection rests on ideas like the "average consumer," required disclosures, and protections for vulnerable buyers, all designed around human judgment Can EU consumer law protect AI agents as algorithmic buyers?. Busch argues that when agents transact with other agents through APIs, there are no screens or mandatory buttons left for those rules to attach to. He proposes moving the obligations into the protocols themselves, for example a withdrawal API that lets you cancel a purchase Can EU consumer law rules survive agent-to-agent API transactions?. A court could find a violation, but it would struggle to enforce a rule written for an interface that no longer exists.

The overall picture: courts are settling narrow questions, such as who accessed a server, while the bigger questions of permission, loyalty and consumer rights are being worked out through platform terms, draft legislation and technical standards. The corpus has only one actual court decision, so it can't say how other courts or other countries will rule. Consumers, meanwhile, are setting their own limits. In one survey, 35% wanted to approve a purchase before an agent completes it Do consumers want to approve AI shopping purchases before they happen?, and in another, most people willing to delegate still wanted to specify brands Do consumers actually want to delegate shopping decisions to AI?.