The longest government shutdown in U.S. history earlier this month disrupted travel plans across America as an air traffic controller shortage led to thousands of flight delays or cancellations. However, even ending the shutdown doesn’t solve the broader air traffic controller deficit issue that has plagued American travelers for years. Artificial intelligence, with its ability to analyze massive amounts of data and potentially transform entire industries, could be helpful. After all, some experts compare AI’s impact on daily life to the invention of electricity or fire. While AI does have some practical applications in aviation, such as processing weather or flight path data, experts have reservations, listing several reasons why this technology is unlikely to replace air traffic controllers anytime soon – and likely never. “It’s a human system, and it’s certainly human-centric, requiring near-flawless performance,” stated Ross Sagun, a former pilot and air traffic controller with over 50 years of flight experience. “The smallest mistakes can turn into gigantic catastrophes.” Why AI won’t replace controllers Air traffic controllers play a critical role in aviation, safely guiding aircraft both on the ground and in the air, ensuring safe separation between aircraft, and communicating with pilots—skills that demand multitasking, quick decision-making, and intense focus. The Federal Aviation Administration (FAA) states it handles over 44,000 flights daily, with 5,500 airborne during peak hours. Nick LaVerne, President of the National Air Traffic Controllers Association (NATCA), described the job as requiring “unique skill and precision” in his written testimony for March hearings before the Subcommittee on Aviation of the U.S. House Committee on Transportation and Infrastructure. He called air traffic controllers the “backbone” of the National Airspace System, adding that they “require rigorous training, mastery of complex systems, and the ability to work under immense pressure.” According to Sagun, it is this pressure that prevents handing this job over to AI. Picking up on emotions and stress in a pilot’s or a fellow controller’s voice is a key signal that something is wrong, and it is unclear if AI can recognize these cues. “That’s where you step in, you hear someone cry for help, and we direct people there to take control of the situation,” he said. The speed of AI development and change would also make certifying the technology difficult, he added. There is virtually no room for error in air traffic control due to the safety risks. Even if the technology were used to automate aircraft management during landing or takeoff, humans would still need to handle unforeseen circumstances, according to Amar Gupta, a research scientist at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). And some research dating back to the early 2000s suggests that a format where a human intervenes in case of an emergency or automation error is not always effective. For instance, a 2005 study found that controllers were more likely to spot incidents manually than when using an automated tool that wasn’t 100% reliable. Air traffic controllers might be less vigilant if automation plays a larger role in their work because they won’t be as actively involved in what is happening, believes Michael McCormick, an associate professor and air traffic management program coordinator at Embry-Riddle Aeronautical University. “This is something people are very bad at—losing vigilance very quickly, usually due to boredom,” McCormick said. How AI can help air traffic controllers Experts believe AI can be useful in air traffic management, but only to assist in duties, not to replace specialists. As many air traffic controllers either retire or take other jobs during the shutdown, such assistance could be timely. Air traffic controller staffing has sharply declined since 2012, according to data cited by NATCA, and the FAA’s Air Traffic Controller Staffing Plan shows the number of certified professional controllers fell below targets in 2024. AI could be helpful in assisting controllers in understanding weather conditions, potentially helping to forecast the likelihood of storms, according to Sagun. The FAA is working on improving a system that uses AI to analyze aircraft trajectories in flight, assess potential conflicts, and provide recommendations to controllers, McCormick revealed, though he notes that it is still several years away from testing and implementation. Searidge Technologies, owned by UK-based air traffic management service provider NATS, has developed an AI system to monitor ground movements and facilitate communication with pilots. This technology is being used at London Heathrow Airport to track aircraft leaving the runway and at Fort Lauderdale-Hollywood International Airport in Florida to monitor gate occupancy. The company states on its website that the technology does not replace human oversight. Ron Reddick, Vice President of Global Sales at Searidge Technologies, told CNN via email that they are seeing “significant global interest” in their system and expect more “operational rollouts and demonstration sites next year.” But don’t expect major upgrades in the United States overnight. Implementing new technology in air traffic control requires extensive testing and is a long process based on decades-old infrastructure. The Department of Transportation had planned to upgrade the FAA’s air traffic control system by 2028, implementing “all new equipment” and “all new software,” Secretary Pete Buttigieg said in May. However, Buttigieg recently told reporters at a press conference that these updates will now take longer than originally planned. “The whole process takes years,” McCormick responded when asked how long it would take to integrate AI into existing air traffic control systems. “We are in the early stages. We haven’t gone very far yet.”
About the Author
