News · 2026-11-12

United Airlines bans unauthorized AI agents from booking flights, even with the traveler's consent

United Airlines said on Wednesday, November 11, 2026, that it is banning unauthorized third-party AI agents from making flight reservations on behalf of travelers. CBS News reports the airline cited security risks and possible violations of federal rules, and the restriction applies both inside United's digital platforms and "indirectly via travel agencies and aggregators".

Key facts

• What: United bars "unauthorized, third-party personal AI agents or assistants" from booking; it is also updating its booking, ticketing and distribution policies.
• When: announced November 11, 2026; bookings made by agents before then will reportedly be honored.
• Who said what: United's statement, CEO Scott Kirby ("I don't want to create another middleman") and analyst Henry Harteveldt of Atmosphere Research Group.
• Primary coverage: CBS News, Business Traveller and Simple Flying.

What United says and what the rule covers

United's stated reasoning is that 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." It gave a federal-rules example: flight check-in requires a human to acknowledge that bags do not contain banned substances like explosives, so an agent doing check-in could violate FAA and DOT regulations. Simple Flying ties that to 49 CFR 175.25, a legal reading that is the publication's own, not United's.

According to Business Traveller, the revised Contract of Carriage says an unauthorized assistant cannot make, modify or service a ticket or buy other products, and covers agents built into browsers. PhocusWire adds that United's Digital Trust Center says travel agencies and booking platforms cannot use third-party AI agents to book United travel and may use their own AI tools only with United's permission. United told PhocusWire it has no partnerships with third-party AI agents today. We could not open United's contract text, so rule numbers cited by Simple Flying are unverified.

Motive: control of the sale

Security is the stated reason, but the analysts point elsewhere. Harteveldt told CBS: "United is saying they don't want to add yet another, and potentially even more powerful, intermediary than they already face", meaning Google search, metasearch and online travel agencies. He also said flight searches cost airlines money, so a surge of bot searches is an expense. Kirby, per the Wall Street Journal as relayed by Business Traveller, said: "To be honest, I don't want to create another middleman." We could not open the WSJ original.

The detail that separates policy from pure security is that consent does not matter. A booking is barred even if the traveler personally told the agent to make it and watched each step. And because an agency booking does not reveal who used the agency's site, blogger Gary Leff argues that case is the hardest for United to police; in practice the rule may rest on United's right to cancel.

Two models for agent commerce

Southwest took the opposite approach. Its October 6 press release announced a plugin in ChatGPT, built with OpenAI and AWS, saying it lets the airline "appear earlier in the Customer journey" with a path "from inspiration and discovery to booking options with Southwest." CBS says Southwest made its inventory viewable to ChatGPT users, who access the airline's website through the app. United's ban is on unauthorized agents, which leaves room for sanctioned partners, though that inference is ours and no source states it. Simple Flying calls United the first major US carrier to formally announce such restrictions and says Delta was reported last month to be blocking similar tools; we did not verify the Delta claim.

Why it matters

For people building AI agents, this is the first big consumer brand to write the rule into its contract: an agent acting for a user is not automatically a user. The credentials problem is real, which is why scoped agent credentials matter, but a sanctioned, scoped integration is exactly what United's policy does not yet offer.

Caveats

Hacker News numbers varied between our sources (one snapshot showed 569 points and 593 comments on the CBS link), and we found no data on how many United bookings come from agents. United fares remain bookable through Kayak, Hopper and Google Flights per Harteveldt, who adds that blocking bots won't make United's fares any easier to find. No separate effective date was given.

Primary source, verified: read the paper →

Ground Truth. (2026, November 12). United Airlines bans unauthorized AI agents from booking flights, even with the traveler's consent. Ground Truth. https://groundtruth.day/news/united-airlines-bans-unauthorized-ai-agents-from-booking-flights.html

@misc{groundtruth:united-airlines-bans-unauthorized-ai-agents-from-booking-flights,
  title  = {United Airlines bans unauthorized AI agents from booking flights, even with the traveler's consent},
  author = {{Ground Truth}},
  year   = {2026},
  month  = {nov},
  url    = {https://groundtruth.day/news/united-airlines-bans-unauthorized-ai-agents-from-booking-flights.html}
}

Topics: united-airlines · ai-agents · policy · travel · agent-commerce · ai-security