United Airlines to block third-party AI agents from booking platforms

United Airlines announced on November 11, 2026 that it will block unauthorized third-party artificial intelligence (AI) agents from accessing its digital booking platform, the first restriction of its kind by a major carrier.

The airline said AI agents pose an unacceptable risk on two fronts. They could expose sensitive passenger information, and they could take actions that result in unintended purchases or unwanted changes to customer accounts, including within MileagePlus, United's loyalty program.

United also raised a regulatory concern that goes beyond data privacy. In a statement, the carrier said AI agents managing passenger bookings may circumvent Federal Aviation Administration (FAA) and Department of Transportation (DOT) security rules that require human acknowledgement for certain actions. The airline cited one example directly: declarations that checked bags contain no banned or dangerous materials, a step that presumes a person, not software, is answering.

What does the restriction cover?

The scope extends past United's own website and app. The ban also applies to AI agents operating through indirect channels — online travel agents (OTAs) and metasearch aggregators that connect to United's inventory.

United said it will update its conditions of carriage, the contractual terms governing passengers, to reflect the decision. That change would give the airline a formal basis to refuse or unwind transactions executed by third-party AI agents.

The carrier framed the move as consistent with practice in other industries that handle highly sensitive customer data, pointing to banking and retail as sectors that have adopted comparable safeguards.

Why it matters for distribution

The decision lands in the middle of an unresolved industry debate over how AI fits into travel distribution, and what it means for the structure and business model of airline retailing.

Agentic AI — software that can search, book and modify travel on a customer's behalf — threatens a core assumption of the current model: that airlines control the digital storefront and the customer relationship attached to it. If agents mediate the transaction, carriers risk losing the direct connection, the accompanying data, and the ancillary sales that flow through it.

The issue has moved up the industry agenda. At the World Aviation Festival 2026, conversations around the use of AI in commercial aviation took center stage, reflecting how quickly the question has shifted from experimentation to governance.

What does this change for passengers and partners?

For passengers, little changes immediately when booking directly; the restriction targets software acting without authorization rather than travelers themselves. The practical consequences will fall on third parties — AI agent developers, OTAs and aggregators — whose products may lose the ability to complete or manage United bookings unless they reach compliant arrangements with the airline.

For the wider industry, the precedent is the substance of the story. One of the largest airlines in the world has now declared that unsupervised AI access to booking systems is a security and compliance risk, not merely a commercial question.

The move could prove a watershed. Other carriers handling comparable volumes of passenger data, and subject to the same FAA and DOT acknowledgement requirements, may seek to emulate United's restriction, potentially fragmenting the distribution channels that AI travel agents depend on before that market matures.