US Flight Ground Stop: Did the FAA's New SMART AI Tool Play a Role?

November 14, 2026 · Cortex Solutions

A nationwide US ground stop on November 13 came three days after the FAA expanded its SMART AI tool to all 22 centers. What officials said and what it teaches businesses.

On Friday, November 13, 2026, departures at airports across the United States came to a halt after communication failures hit air traffic control centers. After hours of disruption, the US Federal Aviation Administration (FAA) lifted the ground stop that same evening, but one question kept dominating the headlines: did the new artificial intelligence tool the agency had expanded just three days earlier play any part in what happened?

What happened on Friday?

According to a report by US station WMUR, part of the Hearst group, updated on the night of November 13, a nationwide ground stop was issued after communication problems at control centers, and audio from Atlanta captured a controller saying he had been told to stop everything. The report notes this was the first nationwide ground stop in the United States since January 2023, when the NOTAM pilot-notice system failed.

Latin Times, in a report published on the evening of November 13, said the stop covered all departing flights from US airports and that the failure affected the radio communication systems pilots and controllers use to talk to each other.

What is SMART?

SMART stands for Strategic Management of Airspace, Routes and Trajectories. According to WMUR, it is designed to analyze weather, flight paths, air traffic and controller staffing in order to predict congestion and recommend more efficient routes. It is developed by Air Space Intelligence under an FAA contract worth 875 million dollars over 12 years.

According to the same report, the tool was first used in the Washington, D.C. area in September and was then expanded on November 10 to all 22 air traffic control centers, just three days before the ground stop.

Latin Times explains that the tool makes recommendations rather than acting on its own: FAA specialists in Washington review them before they are passed to the Air Traffic Control System Command Center, and the leadership of each center then decides whether to accept or decline them. The FAA says, according to the same outlet, that the tool cannot replace human controllers and cannot take control of any aircraft.

What did officials say?

• US Transportation Secretary Sean Duffy linked the outage, according to WMUR, to aging air traffic control infrastructure including an old analog TDM system, said outdated switches, routers and copper lines were overwhelmed with data, and called on Congress to allocate 50 billion dollars to upgrade the system.
• Homeland Security Secretary Markwayne Mullin said, according to KPRC Click2Houston, that his department is working with other federal agencies, through the cybersecurity agency CISA, to determine whether the new technology contributed to the outage, and described it as a telecommunications issue that poses no threat to the public. Latin Times quoted him as saying: You can't rule out anything at this point.
• WMUR reported that Mullin said there was no indication of a malicious attack, while not ruling out other possibilities.
• The developer, Air Space Intelligence, posted on X that the ground stop was not based on or a result of the SMART system, any of its products or its integration with the FAA, a statement also reported by NBC News on November 13.

What we still do not know

At the time of writing, neither the FAA nor the Department of Homeland Security has identified a final cause. KPRC reports that officials have not established that the AI system caused the problem and that FAA technicians are still investigating the communications failure. Latin Times adds that a definitive answer awaits the agency's technical forensics.

The outlet also noted that more than two dozen lawmakers wrote in October to the Transportation Secretary and the FAA Administrator saying the workforce had little to no input into the tool's development, and that controllers had raised concerns it could produce false safety-critical information.

The investigation may ultimately show that the tool had nothing to do with the outage and that the cause was old equipment, as the Transportation Secretary says. But the timing of the expansion made AI the first suspect in the public's eyes, and that alone is an important lesson.

What this means for Saudi businesses

There is no direct link between the incident and the Saudi market, but the management lesson is clear for any organization bringing AI into its operations:

• Clear logs protect the system and its owners: when a new system launches at the same time as any outage, it will be the first thing questioned. Systems that keep a record of what they recommended and when make it easier to clear them or find the fault quickly.
• Keep a human in the loop: SMART's design, based on recommendations that people review before execution, is a useful model for companies, especially for financial or customer-facing decisions.
• Infrastructure first: if the Transportation Secretary's account is right, the problem was old equipment, not the algorithm. Any AI project built on top of aging networks and systems inherits their fragility.
• Involve teams from the start: the lawmakers' point about limited workforce input is a reminder that the employees who will use a tool should be consulted before launch, not after.
• Expand in stages with a fallback plan: moving from a limited pilot to full operation needs time for monitoring, and a manual fallback procedure that everyone knows if the system stops.

As many Saudi organizations move to adopt AI, the companies that build trust in their systems through transparency, human review and gradual rollout will be best placed to scale safely, and to defend their systems when the hard questions are asked.