Aircraft cybersecurity research has spent years focused on wireless attack surfaces: spoofed signals, unauthenticated broadcasts, rogue ground stations. A paper presented this week at the USENIX Security Symposium in Baltimore takes a different angle, and arguably a more uncomfortable one for the industry to sit with. Researchers from the University of California San Diego, working with a collaborator from Oberlin College, demonstrated that a small hardware implant, physically connected to an exposed maintenance port for a matter of seconds, can take over communications inside a Boeing 737’s cockpit electronics.
The target is a communications standard called ARINC 429, a decades-old data bus technology that carries information between key avionics components using electrical current across a pair of wires. It was designed in the 1970s for reliability and predictable behavior, not for a hostile environment where an unauthorized device might try to join the conversation. The bus does not authenticate the origin of messages or cryptographically verify that a command came from a legitimate source. That design has been dependable for decades under normal conditions, but it creates a structural weakness if an attacker can physically connect to the wiring.
The research team built a device that acts as a third party on the bus, overriding legitimate transmissions by driving more electrical current than the genuine equipment. Using a testbed built from real Boeing 737 components, they showed the implant could sit between the flight-management computer and the cockpit display unit, intercepting and replacing messages while suppressing any indication that the data had changed. In their proof-of-concept demonstration, the team showed the technique could be used to alter flight-plan information or manipulate data related to weight, balance, and temperature, values that feed directly into takeoff calculations.
It is important to be precise about the conditions this requires. The attack depends on physical access to a maintenance connector, meaningful advance engineering and planning, and a device built specifically for this purpose. The researchers did not demonstrate the technique on a live, in-service aircraft, and pilots retain the ability to override manipulated instructions, provided they detect that something has changed. The team also disclosed the vulnerability to Boeing well before publication and validated their findings further in Boeing’s own lab, a responsible disclosure process that reflects how this kind of research is meant to work.
What makes the finding significant is not that it represents an imminent, easily executed threat, but that it reframes what counts as an aviation cybersecurity concern in the first place. Physical access to an aircraft has traditionally been treated as a safety and security matter handled through access control, badging, and maintenance procedures, not as a cybersecurity risk requiring the same threat modeling applied to networked systems. This research argues that boundary no longer holds. A maintenance port that is externally reachable, even briefly, is functionally an attack surface, and the industry’s existing physical security controls were not built with that framing in mind.
The 737 is one of the most widely flown commercial aircraft in the world, with thousands in active service, which raises the stakes of any structural finding involving its core avionics architecture, even one requiring the specific conditions this attack does. ARINC 429 is not unique to the 737 either; variations of the same bus technology are used broadly across commercial aviation, meaning the underlying design pattern, not just this one aircraft type, is what deserves attention.
For aviation security teams, the practical response is less about panic and more about updated threat modeling. Maintenance ports, wiring access points, and other physical interfaces to avionics systems warrant the same kind of risk assessment already applied to network endpoints, which means understanding exactly where these access points exist across a fleet, who can reach them and under what circumstances, and whether monitoring exists to detect anomalous behavior on buses that were never designed to expect it. Aviation cybersecurity cannot stop at the edge of the network, and closing this gap requires treating brief physical access with the same seriousness as a remote intrusion attempt.
Source: USENIX Security Symposium ‘26 / UC San Diego - Researchers Use a Physical Device to Take Over Electronics in a Boeing 737
https://today.ucsd.edu/story/researchers-use-a-physical-device-to-take-over-electronics-in-a-boeing-737