United Airlines bans outside AI agents from booking flights

Why is United blocking AI assistants?

United gives three reasons. First, sharing account credentials with an AI assistant could leak sensitive information. Second, an AI tool could make purchases or account changes nobody intended, or spend miles from the MileagePlus loyalty program. Third, automated operation could break rules from the FAA and the US Department of Transportation. The example United gave is check-in, where a human has to confirm that a bag holds no banned items such as explosives.

United plans to update its Digital Trust Center, its security information page. It will explain how the limits affect customers, how to protect an account, and where to go if an AI agent does something it shouldn't. The company also said bank and retail leaders have taken similar steps, and urged other firms to consider the same.

Part of a wider push against Meta's Muse

United is not alone. Meta released its personal AI agent Muse in September, and outside agents have been shut out of one industry after another since. Amazon blocked shopping by Muse from September 20, citing no advance notice from Meta, Muse not identifying itself as an AI agent, and signs that it collects and stores customers' login details.

In banking, JPMorgan Chase, Charles Schwab, Citigroup and Wells Fargo acted between October 13 and 20. They either blocked outside agents that log in with a customer's own credentials, or shifted the risk to the customer by contract. JPMorgan CEO Jamie Dimon said on an earnings call that the bank would not let agents such as Muse or Instinct operate inside customer accounts, citing worries about "rogue agents". Bank of America emailed customers on October 27 telling them not to share credentials with AI assistants, and warned that customers are liable for transactions that result.

For travelers, the practical point is simple. Until United says otherwise, booking and check-in should go through your own login, not an AI assistant.