United Airlines Blocks AI Agents from Booking Flights Over

United Airlines Blocks AI Agents from Booking Flights Over Security Attestation Gaps


Why Your AI Assistant Can't Complete a Booking


United Airlines has drawn a firm line: AI agents cannot book flights on its platform. The reason is not a technical limitation but a legal and security one. When a passenger books a ticket, they must attest that their baggage does not contain explosives. An AI agent cannot make that attestation on a human's behalf—at least not in any way that satisfies current aviation security regulations.


The Attestation Problem


The core issue is the legally binding declaration every passenger must sign before flying. This attestation—confirming no prohibited items, including explosives, are in checked or carry-on luggage—is a personal affirmation required under federal aviation security rules. As of 2026, no regulatory framework exists that allows an autonomous AI system to make this declaration for a traveler. United's position reflects that gap: until an AI agent can legally attest to baggage contents, it cannot complete a booking end-to-end.


Industry-Wide Implications


United's stance highlights a broader tension in the travel industry as AI booking assistants become more capable. While these agents can search fares, compare itineraries, and even fill out passenger details, the final compliance step remains a human responsibility. Airlines and regulators are still working out how—or whether—AI agents can be delegated this authority. For now, travelers using AI assistants will need to complete the final booking step themselves, ensuring the security attestation is made by a human.


What This Means for Travelers


If you're using an AI agent to plan a trip, expect a handoff at the final booking stage. United's policy is a reminder that certain legal and security obligations remain firmly human. As AI capabilities expand in 2026, the question of how to delegate legal attestations to autonomous systems will likely become a larger regulatory debate—but for now, the answer from United is clear: your AI agent can plan, but it can't promise you're not carrying explosives.

via The Verge AI

Related