# United blocks personal AI assistants from booking flights, drawing a line on automated travel

Published 2026-11-11T22:23:46+00:00 · Updated 2026-11-11T22:23:47+00:00 · By superintelligencenews@gmail.com
Canonical: https://superintelligencenews.com/applications/ai-agents-united-blocks-flight-booking/

> United Airlines will block personal AI assistants from completing flight bookings on its site, citing privacy and regulatory concerns. The airline says agents may still show flights, but not finish transactions.

## Key facts

- United Airlines said on Nov. 11 it will block personal AI assistants from booking flights on its site.
- The Verge reported that AI agents may still display United flights but cannot finish the transaction.
- Scott Kirby said on Tuesday: “To be honest, I don’t want to create another middleman”
- United says it has not seen many customers try to use AI agents for flight booking yet.
- The company says it may revisit the policy as third-party AI agents improve.
- The Verge said other companies restricting unauthorized AI agents include eBay and some major banks.

United Airlines will not let personal AI assistants complete ticket purchases on its booking site, a move that puts one of the world’s biggest carriers on the front edge of a fight over who gets to act for travelers online. The airline says the restriction is meant to address privacy and regulatory concerns, and it could limit how useful newer assistants from companies such as Meta and OpenAI are when customers try to book flights.

The policy, reported by The Verge on Nov. 11, stops short of banning AI tools from showing United flight options. But it prevents those assistants from finishing a transaction on the airline’s site, a distinction that matters because it draws a hard line between research and purchase in the emerging market for agentic software.

## What United is blocking

United’s new stance is specific: AI agents may still surface United flights to users, but they cannot complete the booking. According to The Verge, the airline will not allow personal AI assistants to finish ticket purchases on its booking site.

That boundary matters because many of these tools are built to act on a user’s behalf across the internet. In travel, that can mean searching routes, comparing fares and then moving a customer all the way through checkout. United is now saying the last step is off limits unless the system meets the airline’s standards.

### Why does United say it is drawing the line?

United says the problem is not simply automation. It is the combination of sensitive customer data, possible unintended purchases and the possibility that an outside agent could step around airline-required disclosures.

Among the concerns cited by The Verge are privacy risks from agents exposing customer information, the chance that an assistant could make a purchase the user did not mean to authorize, and the risk that automated systems might bypass mandatory attestations. United pointed to an example involving an airline-required explosive-bag attestation.

The airline also says direct interaction between third-party agents and accounts or bookings could involve sensitive actions, and that protections for travelers need to be established first.

## Why it matters

This is more than a company policy tweak. It is an early sign that major consumer platforms may not automatically welcome AI agents that try to move from recommendation into execution.

For travelers, the immediate effect is practical: the newest personal assistants may still be able to help plan a trip, but they may become less useful at the moment that matters most — booking the ticket. If the booking path is blocked, users may need to switch back to a direct airline site, reducing the appeal of an assistant that promises to handle the whole transaction.

For airlines, the issue is control. Ticket sales are not only commercial transactions but also compliance-heavy processes, with disclosure steps and account safeguards. A system that can click through those steps without a human directly present raises questions about consent, customer protections and liability.

For the AI companies building consumer assistants, United’s move is a warning that access to high-value services may depend on negotiation, permissions and technical guardrails rather than broad default access. The policy could reduce the usefulness of newer personal AI assistants from Meta and [OpenAI](https://openai.com/) if those tools cannot complete bookings across major travel sites.

## How the dispute fits into a broader shift

The airline’s position is part of a wider pattern in which companies are trying to control how outside software interacts with their digital services. The Verge noted that other businesses have also restricted unauthorized AI agents, including eBay and some major banks.

That matters because the question is no longer whether AI agents can perform tasks online. They already can, if websites permit them. The deeper issue is whether those tasks can include purchases, account access and other actions that were designed around direct human use.

In travel, the stakes are especially high. Booking systems often require a series of confirmations and disclosures before a customer can pay and travel. If an assistant is acting on the customer’s behalf, the airline still has to know the right person understood the rules and agreed to them. That is the problem United says it wants solved before it opens the door.

## Background: how we got here

AI agents are software systems that can act for users instead of merely answering questions. In practical terms, that means they can search, compare and, where permitted, carry out tasks such as shopping or booking services. The promise is convenience: a user asks, and the system acts.

That promise has been accelerating as companies such as Meta and OpenAI expand their consumer-facing assistants. These tools are increasingly designed not just to chat, but to do things. That evolution has forced online businesses to decide whether they want those systems operating inside their booking flows.

United is not alone in wrestling with that question. According to The Verge, the airline said it has not seen many customers try to use AI agents for flight booking yet. The Wall Street Journal reported the company’s limited observed usage and the policy change context.

That detail is important. The policy is not a reaction to a massive wave of agentic booking. Instead, it appears to be a preemptive move, shaped by what United sees as the risks of allowing a third-party system to act in a transaction that includes personal information and regulatory compliance.

## What Scott Kirby said

United CEO Scott Kirby addressed the issue at an event on Tuesday, making clear that the airline does not want an additional layer between itself and the customer.

> “To be honest, I don’t want to create another middleman”

The quote captures the commercial logic behind the policy. Airlines have long tried to keep control of their direct relationship with passengers, both to manage service and to avoid handing margin and customer data to intermediaries. AI agents, at least in their current form, can resemble a new kind of intermediary — one that can search and transact on behalf of the user.

United also said in a statement:

> “As third-party AI agents advance, we see the opportunities they present.”

That line leaves open the possibility that the policy is not permanent. United says it may revisit the restriction as third-party AI agents improve, suggesting the airline is willing to adapt if technical safeguards and verification methods catch up with the capabilities of the tools.

## By the numbers

| Item | Detail | Why it matters |
| --- | --- | --- |
| Nov. 11, 2026 | The Verge published the report on United’s policy | Marks the public disclosure of the airline’s stance |
| Tuesday | Scott Kirby made his remarks at an event | Shows the policy had executive backing |
| 1 | Specific booking channel targeted: United’s booking site | Limits the restriction to direct transaction handling |
| Several | External companies already restricting unauthorized AI agents, including eBay and some major banks | Places United within a broader industry response |

## What exactly changes for travelers?

For most customers, the near-term effect is likely to be subtle but important. An AI assistant may still help identify United flights, but the final sale cannot be completed by that assistant on United’s site.

That means travelers may still use a chatbot or personal assistant for inspiration and comparison, but they may need to take over once they reach checkout. In other words, the assistant can accompany the search, but not the purchase.

That distinction could frustrate users who expect a seamless experience from a single AI tool. It also places a practical limit on the kind of end-to-end travel automation that companies have been promoting as the next step in consumer software.

## How are airlines and AI companies likely to respond?

The most immediate response may be technical. The Verge said United has not yet seen much consumer use of AI agents for booking, which suggests the airline is acting before the behavior becomes widespread. The open question is how it will identify and block unauthorized agents without interfering with normal customer access.

Another open question is whether other airlines will adopt similar policies. If they do, AI travel assistants could find themselves able to search widely but able to buy only on a patchwork of terms that vary by carrier.

AI companies, meanwhile, will likely need to prove that their agents can respect site rules, preserve user consent and avoid triggering the kinds of concerns United raised. Until that happens, carriers may keep the last click for themselves.

## What happens next

The immediate next step is not a new deadline but a test of implementation. United will have to enforce the restriction in a way that prevents booking by unauthorized agents while still allowing ordinary customers to search and buy tickets directly.

The broader next phase is a standards question. If airlines, banks and marketplaces continue limiting outside agents, pressure may build for clearer rules on how AI systems can authenticate users, honor disclosures and handle regulated transactions.

For now, United is signaling caution. It sees promise in third-party AI agents, but only after protections are in place. Until then, the airline is keeping booking control where it has always belonged: with the carrier and the traveler, not the software in between.

## Timeline

- **Tuesday:** Scott Kirby says he does not want “another middleman” at an event.
- **Nov. 11, 2026:** The Verge reports United will block personal AI assistants from completing bookings on its site.
- **Future:** United says it may revisit the policy as third-party AI agents improve.

## Why this policy may spread

United’s move is likely to resonate beyond aviation because it addresses a problem many digital businesses now face: how much autonomy to grant software that looks like a user but behaves like a service layer.

For companies that sell directly to consumers, an AI agent can be both useful and threatening. It can increase convenience, but it can also reroute control, data and revenue. That tension is already visible in retail and finance, where the need to protect customer information and transaction integrity is especially acute.

United’s decision shows that the first wave of AI agent adoption is likely to be negotiated site by site, sector by sector. The technology may be advancing quickly, but the rules around it are being written much more slowly.

The airline is not shutting the door on the category altogether. It is telling the industry that before software can book a flight, the industry must decide what that software is, what it can see, and who is responsible when it acts.

## Frequently asked questions

### Why is United blocking AI agents from booking flights?

United is blocking AI agents from completing bookings because it says they create privacy and regulatory risks. The airline cited possible exposure of customer information, unintended purchases and the possibility that automated tools could bypass required travel disclosures.

### Can AI assistants still show United flights?

Yes. United’s policy does not stop AI assistants from displaying flight options, according to The Verge. The restriction applies to completing the purchase, not to searching or showing itineraries.

### Will this affect assistants from Meta and OpenAI?

Yes, potentially. The Verge said the policy could reduce the usefulness of newer personal AI assistants from Meta and OpenAI if they cannot complete booking transactions on United’s site.

### Is United the only company restricting AI agents?

No. The Verge said other companies are also restricting unauthorized AI agents, including eBay and some major banks. United’s move fits a broader trend of companies trying to control automated access to their services.

### Could United change the policy later?

Yes. United says it may revisit the policy as third-party AI agents improve. The airline has said it wants protections for travelers established first, which leaves room for a future update if those safeguards emerge.

## Sources

- [The Verge](https://www.theverge.com/ai-artificial-intelligence/1031623/united-unauthorized-ai-agents)