United Airlines announced that it will restrict AI agents from completing bookings on its website, third-party travel sites, and ticket aggregation platforms. This measure aims to reduce the risk of sensitive data leaks, such as payment information and login credentials, and to prevent issues like "accidental purchases," account changes, or MileagePlus points misuse. This decision aligns with similar moves by companies like JPMorgan Chase and Amazon, which have chosen to limit booking processes to human agents in consideration of user protection and customer relationships, and complies with current regulations set by the Department of Transportation and the Federal Aviation Administration.
The U.S. airline industry is facing ethical and regulatory challenges regarding "agent business." United pointed out that allowing AI agents to complete ticket bookings could trigger existing regulatory restrictions; for example, important safety confirmations during checked baggage inspections must be made by humans. CEO Scott Kirby has also publicly expressed a desire to avoid adding another intermediary, reflecting the company's pursuit of balance between customer relationships and operational control.
At the same time, there are differing opinions on the actual acceptance of agent business. In the competitive landscape of airline technology, Southwest has opened a ChatGPT plugin, allowing users to complete part of the booking process directly through the bot; this move indicates diverging strategies among companies. Observing the overall market, while most companies choose to impose restrictions, the popularity of agent-style purchases among the general public remains to be seen.
Comparative Perspectives on In-Flight Entertainment and Automation Technologies in Consumer Experience
Using Uber as a case study, it disclosed during its recent earnings call that it completed 320 million rides and deliveries during the "Halloween week" of the third quarter; among these, AI agents completed 500 Uber Eats orders, while only generating 4 ride services. These figures indicate that "robotic services" have not yet achieved widespread adoption in practical use. This phenomenon suggests that for airlines like United, the risks of allowing AI automated interfaces to intervene in the booking process still outweigh the benefits, particularly in sensitive areas involving financial and personal data.
On the other hand, in-flight streaming and automation support are enhancing passenger entertainment and operational convenience. Even if not directly related to bookings, the application of AI and digital services under in-flight internet can also promote customer loyalty to brands, as passengers have a higher demand for immediate, seamless services during long flights. This trend compels airlines to balance "convenience" with "safety," avoiding risks introduced by automation that could undermine overall customer trust.
Practical Considerations of Privacy, Risks, and Customer Relationship Management
While this decision focuses on the booking process, it also reflects a common issue faced by many large organizations: how to provide convenience while avoiding risks of data leaks and misuse. For instance, in the Muse incident, Meta faced criticism over privacy issues related to user addresses caused by mechanical agent services. These cases remind travelers and businesses that during the use of automation and robotic interfaces, the ultimate control over data remains with the user.
In this context, United and other industry participants also need to implement stricter identity verification, minimum data principles, and clear risk disclosure mechanisms to ensure users are aware of how their data will be used and stored. This article also calls on airports, airlines, and technology providers to offer clear terms of use and transparency before promoting agent interfaces, to avoid undermining customer trust in the brand.

