United Restricts Unauthorised Third-Party AI Agents

United Airlines is restricting unauthorised third-party AI agents from using its digital platforms, and from reaching those systems through travel agencies and aggregators. The carrier set the policy out on 11 November 2026 in a Newsroom release and in an updated page on its Digital Trust Center.

The company says the work includes updates to its booking, ticketing and distribution policies. Bookings already made by an agent before 11 November will not be cancelled. Assistive technology for customers with disabilities is not affected. United also says travellers may still use AI to research and plan a trip, as long as the tool does not access united.com or the United app.

Credential sharing, check-in, and what United is blocking

“sharing account credentials with personal, third-party, AI assistants can expose sensitive information, and actions taken by these tools may result in unintended purchases, account changes, or use of MileagePlus miles,” the release says. It goes on: “these automated AI actions have the potential to violate FAA and DOT regulations – as an example, flight check-in requires human acknowledgment that bags do not contain banned substances like explosives.”

“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.”

The Trust Center is more specific about what counts as unauthorised. If a third-party AI agent, bot, browser extension or browser feature searches, books, buys, manages or collects information on united.com or the United app for a traveller, and United has not given permission, it is unauthorised. That remains so even if the traveller has themselves granted the tool access. “Just because you’ve given access to AI assistants doesn’t mean United will permit an AI assistant to transact with United,” the page says.

On whether United already works with any third-party AI agents, the Trust Center’s answer is “Not yet.” Travel agencies and booking platforms, it adds, cannot use third-party AI agents to search for, book or manage United travel for their customers on direct or indirect channels, and may only use their own third-party AI tools with United’s permission.

What still works, and what United wants travellers to check

The exceptions are as important as the restriction. Assistive technology for disabled customers is carved out. Trip research and planning are still allowed if the tool stays off United’s website and app. The company even notes that some AI tools “including ones built by us” can help with inspiration, while telling customers to stick to general information and not to provide sensitive or private data.

Existing agent bookings are treated as a cut-off, not a purge. “agent bookings made prior to November 11, 2026 will not be canceled,” the Trust Center says. It still asks customers to review recent travel plans to check they have been booked and ticketed correctly. Anyone who has already shared MileagePlus login details with a third-party AI agent is told to change their password immediately.

That advice sits close to guidance TSN has already covered. Singapore’s four-step safety advice for personal AI agents likewise treats credential sharing as a first-order risk. TSN’s own explainer on how to book a flight with Grok Bot and approve the fare yourself is built around a human approval step; under United’s definition, a tool that then goes on to search or book on united.com or the app would still be unauthorised unless the airline had given permission. Keeping logins out of a chat, including in a dedicated vault for saved credentials, is one way to avoid the password-sharing United is warning against.

A line between planning a trip and transacting on United’s systems

United is not banning AI from travel. It is drawing a line between using a model to think about a journey and allowing an automated agent to act inside the airline’s own accounts, inventory and check-in flows. The safety example it chooses is not a data leak but a regulatory one: US check-in requires a person to confirm that bags do not contain explosives. If an agent completes that step, the airline says, the traveller may miss safety content, and the action may conflict with Federal Aviation Administration and Department of Transportation rules. That is United’s stated concern, not a finding that any particular tool has already caused a violation.

The commercial concern is simpler. MileagePlus credentials, confirmation numbers and identity details are, in the airline’s telling, too much authority to hand a third-party assistant that can spend money, change an account or burn miles. The Trust Center’s examples of questions it considers safer are general ones about destinations and restaurants; the examples it tells people to avoid include passwords, confirmation numbers, phone numbers and other personal data used to validate identity.

The policy is also a distribution decision. By reaching “indirectly via travel agencies and aggregators,” United is trying to close a side door, not only the front door of its app. Whether agencies can still use their own approved tools depends on permission United has not, on this page, said it has granted to anyone. “Not yet” is an official position about current partnerships, not a promise about future ones; the page says it will be updated if that changes.

What the notices do not do is name the tools being blocked, set out the technical method of the restriction, or give a date by which the policy updates will be complete. They also do not cancel tickets already written before 11 November. The practical tests will be whether a booking or check-in attempted by an unauthorised agent is in fact refused, and whether travellers who only use AI to plan a trip, off United’s own channels, can still do so as the Trust Center describes. Until United names an authorised partner, the default on its platforms is a prohibition.

Sources

1. United Airlines, “United Restricts Access for Third-Party AI Agents,” 11 November 2026. https://united.mediaroom.com/news-releases?item=125496
2. United Airlines, Digital Trust Center, “Unauthorized third-party AI agents or assistants.” https://www.united.com/en/us/fly/help-center/digital-trust-center/artificial-intelligence/third-party-ai-agents.html

Related stories

Singapore Issues Four-Step Safety Advice for AI Agents is the nearest TSN account of official advice on keeping personal AI agents away from passwords and other sensitive account details.

Also relevant:

• How to Book a Flight With Grok Bot and Approve the Fare Yourself
• Grok Bot Adds a Vault to Save Logins, Cards and Contact Details