- Southwest banned humanoid and animal-like robots from all flights in May 2026; United and Delta followed with their own bans, effective July 23, 2026, and all three cover carry-on, checked and cargo.
- The bans cite lithium-ion thermal runaway risk, evacuation obstruction and stowage failure as the core safety concerns.
- Existing FAA regulations contain no provisions for autonomous or remote-controlled machines onboard aircraft, leaving each carrier to write its own interim policy.
A 3.5-foot robot named Stewie bought a seat on a Southwest Airlines flight from Las Vegas to Dallas, and the airline’s response set off a wave of robot bans across U.S. commercial aviation. Southwest acted in May 2026; United Airlines and Delta Air Lines followed with their own policies effective July 23, 2026. What the incidents exposed is harder to fix than a policy update: aviation rules were not written for machines that look like people.
What United and Delta Now Prohibit
Both carriers updated their dangerous items lists on July 23, 2026. United’s policy prohibits any AI-driven or remote-controlled machine that mimics human appearance, movement or speech, with the same restriction applied to animal-like robots, according to the airline. Delta’s restriction is framed as temporary but covers robots designed to resemble or imitate humans or animals in appearance, size, movement or behaviour, according to the airline. Both bans apply regardless of the robot’s size or stated purpose, and cover carry-on bags, checked luggage and cargo.
Stewie, Bebop and the Trigger
Southwest moved first after a string of publicised incidents. Stewie drew wide attention for occupying a passenger seat. A second robot, Bebop, is reported to have caused an hour-long delay on a Southwest flight from Oakland to San Diego. Together, they made clear that commercial aviation’s tightly scheduled, heavily regulated environment had no real protocol for autonomous or remote-controlled machines that don’t fit overhead bins or standard seats.
The Battery Risk
Advanced humanoid robots carry multi-cell lithium-ion battery packs with the same thermal runaway profile that prompted carriers to restrict “smart” luggage unless batteries are removed. The packs in full-scale humanoid machines are larger and more energy-dense than those in consumer electronics. An onboard fire mid-flight leaves crew with limited response options. Small toy robots whose batteries fall within existing size limits are generally still permitted; it is the high-capacity packs in full-scale machines that airlines say cross an unacceptable threshold.
Evacuation and Stowage
Battery risk is not the only concern. A 75-pound robot becomes a projectile in severe turbulence. It will not fit a standard overhead bin. In an emergency evacuation, a large machine blocking an aisle or exit door costs passengers critical seconds. Crew have no established procedure for securing a device whose motors could activate accidentally. These are foreseeable scenarios, not edge cases, and existing safety protocols do not address any of them.
The Regulatory Gap
Delta’s policy states that humanoid robots are not considered assistive devices and do not meet the definition of a trained service animal, language that points directly at the classification problem. Current FAA regulations, written before autonomous machines were a commercial reality, contain no provisions for robots operating independently inside an aircraft. The frameworks govern human passengers, their personal items and traditional cargo. None of that maps onto a machine with its own motors, battery system and potential for unintended movement.
The result is a patchwork: three carriers, three separately drafted policies, no unified standard. How the FAA addresses this, whether through emergency rulemaking, updated guidance or a formal docket, is an open question. The liability questions alone, covering onboard incidents, data security and interaction with aircraft systems, have no current regulatory home. Airlines are filling that void themselves, for now.
