United Airlines bans AI agents: bots can no longer book your flights

What's happening

United Airlines, one of the largest carriers in the US, plans to stop third-party AI agents from booking flights for travelers or accessing their accounts. AI agents are programs that do more than answer questions. They carry out tasks on their own, like comparing prices, filling in forms and paying. The airline announced the decision on Wednesday, November 11, 2026. Bloomberg reports that the timing is deliberate, coming just ahead of the Thanksgiving and Christmas travel season.

To make this happen, United is changing its rules for booking, ticketing and distribution. The ban won't stop at its own website and app. It is also meant to apply indirectly through travel agencies and aggregators, the portals that pull together offers from many airlines. The rules explicitly target "unauthorized" personal AI assistants. The sources don't say whether United plans to authorize specific providers in the future.

The official reason: security and regulations

To book a flight, an AI agent needs access to personal data, meaning your login credentials and payment details. That's exactly what United sees as the problem. The airline's statement, quoted by CBS News, reads:

"Sharing account credentials with third-party personal AI assistants can expose sensitive information, and actions taken by these tools can result in unintended purchases, account changes or the use of MileagePlus miles."

MileagePlus is United's frequent flyer program. If an agent redeems miles on its own, the customer loses real value.

The second argument is a legal one. United says an automated check-in by an AI could violate federal rules. At check-in, passengers have to confirm themselves that their bags contain no prohibited or hazardous materials. If a bot clicks through that confirmation, strictly speaking the person never gave it. According to Yahoo Finance, United points to rules from the Federal Aviation Administration (FAA) and the US Department of Transportation.

"That's why United is taking steps to restrict access, directly on our digital platforms and indirectly through travel agencies and aggregators, to prevent use by unauthorized third-party personal AI agents or assistants."

The real reason? Control over the customer

Observers see business interests as the main driver. United CEO Scott Kirby was blunt with the Wall Street Journal, according to Yahoo Finance:

"Frankly, I don't want to create another middleman."

Aviation analyst Henry Harteveldt of Atmosphere Research Group read it the same way in comments to CBS News. On top of Google, metasearch engines and online travel agencies, United doesn't want yet another intermediary, possibly an even more powerful one, sitting between it and its customers. Then there's cost. Every flight search costs airlines money, because fare and availability queries run through expensive booking systems. A flood of AI bots firing off searches en masse would drive those costs up.

Harteveldt says not much should change for customers at first. United fares will still be bookable through comparison sites like Kayak, Hopper and Google Flights, so good deals shouldn't get harder to find. "But by blocking the AI bots, United isn't making it any easier to find fares either," he added.

How other companies handle AI agents

The industry is split:

CBS News names Meta's Muse and OpenAI's dots as examples of popular AI agents. Consumers are increasingly using them to research products, hunt for deals and make purchases. According to the New York Times, the major AI companies have recently been releasing more agents like these, which handle things like restaurant reservations or shopping.

Background: AI in travel and the fight over fares

The demand is there. According to booking platform Hopper, about 33 percent of adults have already used generative AI to research a trip, and Hopper expects that number to climb this holiday season. Domestic airfares have also risen sharply, Hopper says: up 31 percent for Thanksgiving and 23 percent for Christmas compared with last year.

The debate over so-called surveillance pricing makes the issue even more charged. That's when personal data is used to tailor prices to what an individual customer is likely willing to pay. According to Yahoo Finance, the US Federal Trade Commission (FTC) ordered United, Delta, JetBlue and six other major airlines in October to disclose how AI tools affect their ticket prices and how they use customer data. Delta drew criticism last year over plans for AI-driven pricing, but denies setting individual prices based on personal data.

Markets took the news in stride. According to Yahoo Finance, United shares dipped slightly on Wednesday afternoon, as did Meta's.

Our take

United's move is an early, unmistakable signal in the growing conflict between AI agents and the platforms they're supposed to work on. The security arguments aren't made up: if you hand a bot your password and credit card, you're taking a real risk. And the hazmat confirmation at check-in really is meant to be a personal declaration.

But the CEO's own words show that business interests matter at least as much. For years, airlines have tried to steer customers away from travel portals and toward their own channels, where they can sell extras and collect data. An AI agent that compares fares neutrally and simply books the cheapest flight undermines exactly that. Critics will likely also point out that the airline is happy to use AI itself, for pricing for example, while denying customers the same technology.

It's still unclear how United plans to enforce the ban technically. AI agents that operate a website the way a human does aren't always easy to tell apart from real users. It's also unclear what "authorized" actually means, and whether official partnerships like Southwest's deal with ChatGPT will follow. For now, if you're flying United, book yourself or use a classic comparison site, not your AI assistant.