Gate closed to bots: United updates carriage rules to halt autonomous AI checkouts
United Airlines has become the first major legacy carrier to draw a hard line against agentic artificial intelligence, updating its binding Contract of Carriage on Wednesday to explicitly ban and block third-party autonomous AI agents, browser bots, and algorithmic assistants from purchasing flight tickets or managing passenger reservations.
The policy update targets the fast-expanding ecosystem of consumer AI tools—including computer-use agents, automated browser extensions, and virtual concierge bots—designed to independently navigate ticketing portals, input passenger information, select seats, and execute payment transactions on behalf of users. Under the revised terms, any ticket generated through unauthorized third-party automated software may be subject to cancellation without refund.
United’s leadership made no secret of the strategic motivation underpinning the technical crackdown. Speaking bluntly to airline industry executives in Denver, United Airlines CEO Scott Kirby framed the prohibition as an essential defense of the carrier’s direct customer relationship. "To be honest, I don't want to create another middleman," Kirby remarked, drawing comparisons to the costly era when global distribution systems and online travel agencies extracted massive fees from airlines.
Aviation safety rules and loyalty fraud drive operational pushback
Beyond commercial positioning, United presented substantive regulatory and cybersecurity rationales for locking out autonomous agents. Federal Aviation Administration (FAA) and Department of Transportation (DOT) safety regulations mandate that individual passengers personally review and acknowledge statutory hazardous materials warnings—specifically restricting lithium-ion battery banks, aerosol containers, and flammable substances—before a boarding pass can be generated.
United argued that automated bots clicking past federal safety acknowledgements violates the regulatory requirement for direct human confirmation. Furthermore, cybersecurity engineers warned that granting third-party AI assistants access to MileagePlus loyalty credentials exposes millions of passenger accounts to credential stuffing attacks, unauthorized award redemptions, and data scraping exploits.
Automated scrapers also inflict tremendous server overhead on global distribution reservation systems, generating millions of rapid-fire lookups that spike infrastructure costs and degrade real-time seat availability feeds for human fliers.
What fliers can still do with AI and how existing reservations are treated
United clarified that travelers are not prohibited from utilizing conversational AI platforms for destination research, flight comparison, or itinerary planning. The boundary is strictly enforced at the transaction boundary: AI platforms may direct users via deep links to United's website, but the actual checkout, seat selection, and credential entry must be executed directly by a human flier.
For passengers who have already completed bookings utilizing early-generation AI travel tools prior to the November 11 policy rollout, United confirmed those itineraries will remain honored and valid. However, the airline cautioned customers to review their reservation details and seat assignments manually inside the official United app to ensure passenger names, date of birth, and Secure Flight passenger data accurately match government-issued identification.
Aviation analysts anticipate other major U.S. carriers, including Delta Air Lines and American Airlines, may soon introduce parallel restrictions to prevent Big Tech platforms from monopolizing the lucrative corporate travel booking funnel.
