The Federal Aviation Administration is about to put an AI advisor into one of the most congested skies in America. According to Ars Technica, officials say the agency’s SMART system could begin helping manage traffic around Washington, D.C.’s three major airports as soon as Monday, September 21—the first step toward a planned nationwide rollout across the roughly 29 million square miles of U.S. national airspace the FAA oversees.
SMART is not a chatbot in the tower. The FAA describes it as using AI models to predict air traffic flows and flag potential conflicts from factors like airline schedules, weather, airport capacity, and airspace conditions. The pitch: give the FAA, airlines, and operators a shared operational picture so they can agree on more efficient routes and departure times, cut fuel burn, improve on-time performance, and recover faster from weather and congestion messes.
Limited launch first—on purpose
Ars reports that U.S. government and industry officials told the Wall Street Journal the D.C.-area debut is the opening move, not the full national flip. Aviation safety and AI governance consultant Philip Mann—who spent 17 years in multiple FAA roles—told Ars that starting narrow is the “right call.” SMART’s risk, he argued, was never any single prediction; it is a national-scale system with AI components carrying more unknowns than anything the FAA has fielded. Every cut in scope shrinks the unknowns.
Airlines were reportedly confused for weeks about how intrusive the tool would be, Politico reported. Anonymous airline officials said nerves eased only after the FAA clarified SMART would not change controller or airline procedures overnight—instead producing “alternative route information” through existing FAA systems. That framing matters: advisory data layer first, not a silent autopilot for separation.
The $875M backbone behind the predictions
SMART sits inside an $875 million, 12-year contract awarded in June to Boston-based Air Space Intelligence. The same deal covers a Flow Management Data and Services system meant to replace the current setup at the FAA’s Air Traffic Control System Command Center in Virginia. Mann has described that flow-management layer as the “backbone,” with SMART as the predictive layer above it.
Air Space Intelligence says its separate Flyways AI platform already helps manage over 40 percent of U.S. air traffic through airline customers such as Alaska Airlines, using a 4D digital twin of national airspace to anticipate traffic and weather. What exact model classes SMART uses—older deterministic rules engines, statistical machine learning, or generative systems—remains unclear publicly. Ars notes the distinction matters: deterministic models repeat; ML models predict from probabilities; generative LLMs can vary and invent. For safety-critical traffic management, that taxonomy is not academic.
Mann also wants clarity on whether the initial rollout stays limited to aircraft at 24,000 feet and above, as the FAA described in June—cruise traffic plus climbs and descents feeding metro airports. The expansion evidence he wants to see: performance under real workload and degraded data, not demo conditions, plus a written answer to who owns the outcome when a prediction is wrong.
Why the FAA is betting on AI now
The SMART trial lands amid a multi-billion-dollar push to replace aging ATC hardware—1980s-era radars, radios, telecom lines, and voice switches. It also lands against a decades-long controller staffing crunch. A December 2025 GAO report cited by Ars said the controller workforce declined about 6 percent over the prior decade. Hiring freezes during past shutdowns, COVID training limits, 2025 probationary layoffs that hit some ATC-systems staff (later partly reversed by a judge), and a 43-day shutdown in late 2025 that Transportation Secretary Sean Duffy said accelerated retirements all compounded the stress.
In May, Ars notes, the FAA reduced its estimated controller need for 2026–2028 by roughly 2,000 people—suggesting leaders hope modernization and tools like SMART can stretch scarce human capacity. The stakes are not theoretical. Investigators cited tower understaffing among causal factors in the January 29, midair collision near Reagan National that killed 67 people.
Geeknewz take
SMART’s D.C. soft launch is the grown-up version of “move fast”: advise first, keep procedures stable, shrink the blast radius. That is the right product instinct for AI near airplanes. The hard questions start after Monday—whether recommendations stay truly optional under schedule pressure, what model class is actually flying in the loop, and who eats the risk when a prediction is wrong in weather that does not care about your digital twin. Watch the evidence gates for expansion harder than the press release.
Source: Ars Technica — FAA tees up $875M AI tool to help manage air traffic congestion (Jeremy Hsu, Sep 18, 2026), citing WSJ/Politico reporting and FAA contract context.