United Airlines becomes the latest enterprise to shut its door to customer bots

United Airlines on Wednesday became the latest enterprise to block customer bots, following similar moves from Walmart, Target and Amazon.  The $59 billion airline, the second largest in the world, said the move was prompted by fears of data leakage, unintended purchases, and the risk of violating federal airline rules.  But industry observers say that the more concerning business risks are the lack of industry standardization around AI coupled with an ever-increasing number of bots from major AI vendors and AI startups. The fear is that supporting the multiple bots that customers use could become an almost infinite drain on IT resources. However, they said, strictly enforced industry standards might persuade some enterprises to rethink their bot resistance.  United’s statement said, “United is taking steps to restrict access – directly within our digital platforms and indirectly via travel agencies and aggregators – to prohibit use by unauthorized, third-party personal AI agents or assistants. This work includes updates to our booking, ticketing and distribution policies.” It pointed out that corporate leaders in the banking and retail industries have taken similar actions regarding AI agents, and said, “we encourage others to consider measures that preserve the data and security of their customers and strengthen safety.” The airline listed a variety of problems with bots that it is trying to avoid, including the risks associated with customers sharing account credentials “with personal, third-party, AI assistants [that] can expose sensitive information, and actions taken by these tools may result in unintended purchases, account changes, or use of MileagePlus miles.” It also spoke of regulatory challenges: “These automated AI actions have the potential to violate FAA [United States Federal Aviation Administration] and DOT [US Department of Transportation] regulations,” such as those stating that flight check-in requires human acknowledgment that bags do not contain banned substances like explosives. However, the airline sent a clear signal to the industry that it was willing to reconsider the ban if there are sufficient changes. “As personal AI assistants evolve, United will continue to evaluate how these tools interact with our services, with a focus on protecting our customers,” the statement said. Separating the humans from the bots One of the bigger enterprise concerns about agents is that, because they are using the customer’s credentials, the bot activity is typically indistinguishable from the human customer’s activity. Analysts say that the cleanest way to address that is to simply assign bots their own credentials. Last month, Walmart got close to that goal, putting in place a program, scheduled to start this month, to try to differentiate bot and human actions. “Today, when a customer uses an AI agent, it signs in using a customer’s account. This means that all purchases – whether made by the customer or the agent – appear together, without any differentiation,” its announcement said. “Agent Profiles, launching in the next month, will offer a new level of transparency for customers. Customers will be able to authorize each AI agent with a dedicated profile within the customer’s account, so the customers will be able to see the agent’s purchases separately from the rest of their purchases.” Target, on the other hand, blocked all but a handful of approved agents in recent changes to its terms and conditions, granting access only to Target’s AI Shopping Assistant, OAI, Google and Microsoft, although it said it may ”add, remove, or otherwise modify the list of Approved Agents at any time and in its sole discretion.” Any other agentic commerce tools, it said, “are unauthorized, expressly prohibited, and may not be used to access, log into, scrape, query, or otherwise interact with or gather data from the Site under any circumstances, whether directly or through API call(s), automated use of user credentials, or other means” and approved agents may not be used in combination with unauthorized ones unless explicitly approved by Target, in writing. Reducing risk Although analysts and consultants agreed that they expect many more enterprises to block external bot access, their opinions were mixed on whether these blocks would meaningfully dilute the value of autonomous agents, especially those making purchases, ordering plane tickets or performing related concierge functions. Frank Dickson, principal analyst at Dickson Research, said, “blocking does not destroy the value of agents. It moves it. Every wall eventually gets a gate. The gate charges admission.” He pointed out that United’s own wording points there: it prohibits unauthorized agents, which leaves room for authorized ones. “Expect agent access to come back through channels the enterprise controls, on the enterprise’s terms,” he said. “Whoever owns the agent owns the customer relationship. Airlines have spent years pulling customers back from third-party travel sites. They are not going to hand that relationship to someone else’s chatbot.” He added that enterprises are just trying to reduce risk. “Enterprises do not fear bots. They fear liability that nobody has assigned,” Dickson said. “Once a rule says which steps a human must perform, which steps an agent may perform, and how each is recorded, a general counsel can sign off on the deployment. Ambiguity is what keeps bots stuck in pilot.” He noted that the complexity of having an unlimited number of different agents visiting is too risky.  “A sea of agents is a sea of requests. Somebody pays for every one,” he said. “Every new agent arrives with its own way of logging in, its own idea of what the customer approved, and its own record of what it did, if it keeps one at all. An airline cannot build a separate door for each of them. Nobody should expect it to. Saying no until the market settles down is a reasonable risk decision. Bots that check prices over and over do not buy more seats.” Noah Kenney, principal consultant at Digital 520, added that there’s no downside to United’s ban.  “United is in a position where there is substantial risk, an uncertain commercial upside, and almost no competitive pressure to accept either,” he said. “In that type of position, they are able to take the approach of waiting for the legal consequences of bot behavior to become more predictable before they allow it.” Limits needed Kenney said one of his biggest concerns is the lack of effective limits on what a bot can do. “One of the biggest mistakes I see in enterprise AI deployments is giving agents access to far more information and systems than their responsibilities actually require,” Kenney said. “There is a fundamental difference between an agent that is instructed to access only one system and an agent that is technically incapable of accessing anything else. We shouldn’t be giving agents broad capabilities and hoping they exercise restraint.” Because of this, Justin Greis, CEO of consulting firm Acceligence, said he expects more companies to impose restrictions on bots while they establish controls on their use, noting, “widespread blocking would reduce agents’ value by preventing them from completing tasks across services.” He added that customer training in bot realities could also be quite helpful. “People need to understand that actions they authorize through an agent can still create obligations for them, even when they never personally click ‘buy,’ ‘send’ or ‘confirm,’” Greis said. “I expect a substantial cultural shift here: just as people may say things online they would never say face to face, agents could create a similar distance between someone’s choices and their sense of responsibility.”