← Home

United Airlines Bans AI Agents From Booking Flights on Behalf of Passengers

United Airlines updated its contract of carriage to explicitly prohibit the use of artificial intelligence agents in booking, modifying, or paying for airline tickets, seats, and add-ons. The new rule, published on November 11, 2026, took effect immediately and applies to any autonomous or semi-autonomous software, including AI assistants built into consumer platforms, AI browsers, and third-party assistants, regardless of whether the passenger or another individual supervises or approves the action.

The measure extends beyond the airline's website and mobile app. It also covers agents that operate indirectly through travel agencies and price aggregation platforms. According to United's Digital Trust Center, agencies and booking platforms cannot use third-party AI agents to search for, book, or manage United travel on behalf of their customers, and may only use their own AI tools with the airline's explicit authorization.

United's public justification combines two strands. On the plane of digital security, the company warned that sharing account credentials with third-party AI assistants could expose sensitive information, including personal data, travel plans, and travel preferences, and result in "unintended purchases, account changes, or use of MileagePlus miles." On the regulatory plane, United stated that Federal Aviation Administration (FAA) and U.S. Department of Transportation (DOT) rules require certain disclosures to be made by the passenger in person — such as confirming that baggage does not contain explosives during check-in — which an AI agent could not legally fulfill.

However, the strategic motivation was revealed by CEO Scott Kirby himself the week before the rule's publication. At an event in Chicago, Kirby told The Wall Street Journal that he did not want to "create another middleman" between United and its customers. The statement captures a fear running through the entire aviation industry: if AI agents begin completing transactions on behalf of passengers, airlines will lose direct control over sales, customer data, and distribution margins. This is precisely the same concern that led Amazon to block Meta's Muse assistant in September 2026, and that prompted JPMorgan Chase to restrict AI agent access to bank accounts during the same period.

The practical impact of the rule warrants two observations. First, United does not prohibit AI assistants from searching or comparing the airline's fares — they simply cannot complete the transaction directly through United's systems or through its agency partnerships. Second, the company said that reservations made by agents before November 11, 2026 will be honored, but that it may cancel reservations made in violation of the policy without notifying the passenger.

United's move stands out in the current industry context. While the airline is closing its doors to unauthorized agents, other carriers are taking different paths: Southwest Airlines launched an official ChatGPT plugin that allows flight reservations directly through OpenAI's platform; easyJet already offers flight bookings and live flight-status updates through AI assistants; and major platforms like Expedia, Booking.com, and TripAdvisor have been developing AI integrations since 2025. This divergence in approach reflects a broader division over how agentic commerce will be structured: companies that want to control the customer experience directly versus those that see AI agents as a necessary distribution channel, even if unwelcome.

What makes United's rule particularly noteworthy is its extraterritorial reach — it applies to agents operating indirectly through third parties, not just to direct website usage. This means an agent connected to a global travel platform that then books a United flight indirectly would also be in violation. United's next step will likely be implementing technical agent-detection mechanisms at the API level, similar to those already in place for scraping bots. The measure raises a fundamental question: in a world where AI agents begin making consumption decisions on behalf of users, who is the true buyer — the human behind it or the software that executed the action — and who sets the rules of commerce?

Sources: The New York Times, Engadget, View from the Wing

✓ Independent sources cross-checked and verified before publishing