United Airlines has announced that it will not permit personal AI agents to book flights on its platform. The carrier cites specific concerns regarding customer privacy and the need to adhere to federal regulatory standards. This move places the airline among a growing number of businesses that are restricting how autonomous artificial intelligence tools interact with their digital services.
According to a report by The Wall Street Journal, the airline has not yet observed a high volume of customers attempting to use AI agents for ticket purchases. This finding aligns with observations made by other major service providers, such as DoorDash, which have also noted limited adoption of these automated booking methods so far.
Despite the current low usage, United is concerned about the potential risks associated with granting AI agents full transactional authority. The company worries that these agents could inadvertently expose sensitive customer information or execute purchases that were not explicitly intended by the traveler. Additionally, there are concerns that automated systems might bypass specific federal rules regarding the information airlines must share with passengers.
One key regulatory requirement involves customers attesting that their baggage does not contain explosives. United argues that AI agents cannot currently provide this personal guarantee on behalf of the traveler. Because an agent cannot physically verify the contents of a bag, the airline prefers to keep the human in the loop for this specific certification process.
Under the new policy, AI agents will still be able to display available United flights and search for options. However, they will not have the capability to finalize the transaction or complete the booking process on the user's behalf. This restriction could limit the overall utility of newer personal AI assistants, including products like Meta's Muse and OpenAI's Dots, which aim to handle complex multi-step tasks for users.
United CEO Scott Kirby addressed the decision during an event held on Tuesday. Speaking to The Wall Street Journal, Kirby stated that he did not want to introduce another layer of intermediation into the booking process. He emphasized the company's desire to maintain a direct and transparent relationship with its travelers without the potential complications introduced by third-party automation.
United is not the only major corporation grappling with the question of whether to allow AI agents to transact on their websites. Other prominent platforms have taken similar precautions. For instance, eBay appears to have implemented measures to block agents from browsing its site, while several major banks have also placed restrictions on how these automated tools can access their systems.
The airline suggests that its stance is not permanent and that it may reconsider its policy as the technology evolves. United indicated that it is watching for future advancements in the field of third-party AI agents.
In a statement, the company acknowledged the opportunities that these agents present. However, it noted that allowing them to interact directly with a United account or booking system involves sensitive actions. The airline expressed a priority to ensure that established protections are in place for travelers before expanding access to these automated tools.
This decision highlights the broader industry challenge of balancing the convenience of AI automation with the need for security and regulatory compliance. As personal AI assistants become more sophisticated, companies across various sectors will continue to evaluate the appropriate level of access these agents should have to their digital infrastructure.
