Field report E-69 / M2 · sheet /operations

Operations & RestructuringFiled from the territory

FAA says $875 million AI system did not cause flight ground stop

The FAA says its $875 million SMART AI platform, expanded to all 22 control centers on Nov. 10, did not cause the subsequent ground stop, but lawmakers demand an audit of safety guardrails.

Grid
E-69 / M2
Surveyed
November 16, 2026
Reading
4 min read
Correspondent
Elena Vasquez
Questions swirl around FAA's new SMART AI system, though officials say no link to ground stop
Fig. E-69 / M2Questions swirl around FAA's new SMART AI system, though officials say no link to ground stop — bernswaelz / pixabay

Field notes

  1. The $875 million SMART platform expanded nationwide on Nov. 10.
  2. Deployment covered all 22 U.S. air traffic control centers, three days before the ground stop.
  3. A Washington-area test rerouted 180 flights, saving 10 minutes and 80 gallons per flight on average.
  4. House Democrats requested a GAO audit in an Oct. 13 letter.
  5. The controllers’ union represents about 14,000 controllers, including roughly 3,000 trainees.

The FAA expanded its $875 million SMART artificial-intelligence platform nationwide on Nov. 10, covering the airspace of all 22 U.S. air traffic control centers three days before a communications-related ground stop disrupted air travel; officials say the system was not responsible.

Government officials and Air Space Intelligence, the startup that designed SMART, said the stop, which began Friday and caused havoc for travelers, was not linked to the system. ASI said the ground stop was not caused by “the SMART system, any product of ASI or our integration with FAA and ATC.”

Homeland Security Secretary Markwayne Mullin said Friday there was no indication that the communications issues had triggered the stop, and senior Transportation Department and FAA officials reiterated that position Saturday. The timing intensified questions about what SMART does, how its benefits and risks compare with the existing human system, and what guardrails apply when its outputs are wrong.

What is SMART, and who retains control?

SMART, short for Strategic Management of Airspace, Routes and Trajectories, was unveiled in September by Transportation Secretary Sean Duffy and FAA Administrator Bryan Bedford as a platform that consolidates 200 data streams, including weather patterns, flight paths, traffic flow and controller staffing metrics.

The platform’s AI-supported engine synthesizes those inputs and gives specialists a comprehensive visualization of aircraft destinations, airspace capacity, congestion and weather risks; managers can then intervene before problems worsen.

At a single site near Washington, specialists examine flight schedules, weather forecasts and hundreds of other data streams. The office sends recommendations to the FAA command center, which relays them to controllers nationwide for consideration.

The government fact sheet says SMART can predict where traffic is heading, identify alternative routes before congestion closes an area, reopen weather-restricted airspace faster and help manage airport arrivals.

How does the system handle weather and congestion?

The FAA’s current practice, according to the fact sheet, blocks weather-affected airspace in large rectangular boxes, even though thunderstorms often form irregular systems with narrow gaps or corridors that aircraft can safely use. SMART instead projects how weather will affect traffic over time, allowing controllers and operators to plan around storms and use available airspace rather than shut down oversized areas.

What has SMART delivered in testing?

The FAA said an initial Washington-area test this fall rerouted 180 flights and saved an average of 10 minutes and 80 gallons of fuel for each flight by avoiding weather or other problems. One Tampa, Florida, to Dallas flight cut 42 minutes from its journey, more than four times the test’s average time saving.

Why are lawmakers seeking an audit?

Before the ground stop, House Democrats sent an Oct. 13 letter to the U.S. Government Accountability Office requesting an audit of SMART’s safety controls and guardrails. They said the aviation safety workforce had little or no input into development and limited training and information during deployment.

The lawmakers also cited possible interference with experienced controllers’ decisions and the risk that AI models could fabricate safety-critical information, or hallucinate, undermining claimed reliability and safety benefits.

Can AI replace air traffic controllers?

SMART does not yet give AI control of aircraft or flight paths; it supplies recommendations that human traffic managers and controllers may accept or reject. The system is designed for managers responsible for the entire U.S. airspace rather than one airport or regional center.

ASI co-founder and CEO Phillip Buckendorf told NPR in October that the technology performs computations while operators retain authority. “They make the decision, they are the decider, they are the judge,” Buckendorf said.

Mike O’Donnell, a former senior FAA official who spent 17 years at the agency and now consults on aviation safety, said the FAA cannot afford to be wrong. He urged officials to treat SMART as one tool rather than the mechanism doing the work, and to verify and validate what it sends controllers.

O’Donnell said the technology should earn repeated proof in operation before receiving greater trust, even with proper guardrails and backstops against incorrect information.

Michael McCormick, a former FAA airspace manager and Embry-Riddle Aeronautical University associate professor, said no other country has a working AI air-traffic tool like SMART. Eurocontrol has a system in testing, while several manufacturers are prototyping similar tools.

The rollout also lands amid a long-running shortage of air traffic controllers. The National Air Traffic Controllers Association has raised the issue with Congress for more than a decade and told lawmakers in 2015 that staffing had reached crisis levels after the FAA missed hiring targets for five consecutive years.

The union represents about 14,000 controllers, including roughly 3,000 trainees. Experts say no AI system can replace controllers, whose demanding, stressful jobs require years of training; SMART instead shifts data synthesis and early problem detection to software while preserving human judgment.

The immediate test is therefore not whether SMART can direct a flight, but whether the FAA can prove that its human-in-the-loop safeguards remain reliable at national scale.

via abc11.com (Source)

Share this article:

More from Elena Vasquez

Elena Vasquez

Show full bio

Staff writer covering companies, deals and strategy in corporate strategy and chief-executive leadership at CEO Frontier.

486 articles

Filed under

Adjacent sheets

« Previous articleNext article »

End of sheetE-69 / M2 · November 16, 2026