AI agents promised to make life easier in many ways. Just give them a task and let them handle all the tedious work for you. Booking flights, dinner reservations, or even dental appointments were a few of the early examples we were promised. But it seems that United Airlines isn’t too thrilled about letting AI handle that final step.
The airline has announced that it’s blocking unauthorized third-party AI assistants from booking flights on behalf of customers. According to The Wall Street Journal, the restrictions target AI agents capable of completing transactions independently, including the emerging category of personal assistants from companies like OpenAI and Meta.
The company apparently has several reasons for doing so, and it’s not just about privacy.
United doesn’t want AI playing travel agent
In an official statement, United explained that allowing third-party AI assistants to access customer accounts introduces significant security risks. Such tools could potentially expose sensitive personal information, make unintended purchases, modify bookings, or even spend MileagePlus rewards without customers realizing it.
There’s also a regulatory complication. Certain airline procedures require direct acknowledgement from passengers, including confirming that checked bags don’t contain prohibited items such as explosives. United argues that allowing AI to perform these actions automatically could create problems with Federal Aviation Administration (FAA) and Department of Transportation (DOT) requirements.
So now, United is updating its booking, ticketing, and distribution policies to restrict unauthorized AI agents across its own digital platforms, travel agencies, and flight aggregators. However, this doesn’t mean AI is completely off-limits for planning your next trip. According to the WSJ, these assistants can still surface United’s flight options and fares, but they won’t be allowed to complete the booking themselves.
The bigger battle over who controls your bookings
Speaking at an event in Chicago, United CEO Scott Kirby told the WSJ, “I don’t want to create another middleman.” This statement highlights a growing tension between companies developing autonomous AI assistants and businesses that would prefer customers to interact directly with their platforms.
United isn’t alone in pushing back against AI agents. Amazon and several major banks have also introduced restrictions on unauthorized AI agents, citing similar privacy and security concerns. Southwest has a similar, albeit slightly different approach, by introducing a ChatGPT integration that lets travelers search for flights. On the other hand, Delta reportedly allows properly authenticated AI assistants to make reservations, with additional safeguards for sensitive bookings.
So there’s definitely a divide in how companies are taking on this new challenge, while brands like Google and OpenAI are letting AI agents handle everything from shopping to travel arrangements.
