More than 2,000 flights were canceled after Britain’s air-traffic system failed for roughly four hours on Tuesday, and a new report now links the breakdown to data submitted for a military aircraft. The Financial Times, citing four people briefed on the matter, reported late Friday that “spurious” flight information triggered the failure. That account offers the clearest explanation yet for a disruption that stranded or delayed hundreds of thousands of travelers, but it is not a final official finding.

The distinction matters. Britain’s Ministry of Defence said there was no indication that the military had made an error, according to a Guardian report, and National Air Traffic Services, known as NATS, said its investigation remained active. The available evidence therefore points to a problem at the boundary between submitted flight data and the system that had to process it; it does not yet establish whether the filing itself was defective, whether a valid but unusual plan exposed a software weakness, or whether both were involved.

What the New Account Adds

The reported military connection changes the focus of the inquiry from a generic technology failure to a specific data-handling event. According to the Financial Times account, information associated with one military aircraft entered NATS’s flight-processing system and precipitated a wider shutdown. The report also said the system did not have a dedicated backup capable of taking over in the same way during a complete failure, a claim likely to intensify scrutiny of how Britain protects a service on which airlines, airports and passengers depend.

NATS has publicly confirmed less. In its incident updates, the company said at 3:25 p.m. Tuesday that engineers had located the issue in the flight-processing system. A fix was in place by 4:40 p.m., and NATS said at 7:30 p.m. that the system was operating normally. It apologized repeatedly and warned that recovery would take time because the interruption had created a backlog across the aviation network.

Chief Executive Martin Rolfe ruled out a cyberattack and said the failure was unlike incidents the organization had previously encountered. That statement is not necessarily inconsistent with a problematic flight plan: a new combination of data and system behavior could produce a different technical failure even if flight information was the initiating event. The investigation will need to separate the triggering input from the system design decisions that allowed one event to spread.

How Four Hours Became Days of Disruption

Airspace did not become unsafe or fully close. NATS restricted the rate at which aircraft could depart while controllers and engineers managed traffic under reduced capacity. Those safeguards protected the operating system, but they rapidly consumed the slack in tightly scheduled airline networks. When departure slots disappeared, aircraft and crews failed to reach the airports where later flights were supposed to begin.

By Wednesday, the immediate fault had been fixed but the network was still unwinding. Reuters reported that about 2,000 flights had been canceled and airlines faced losses measured in millions. A separate AP account described cancellations across at least 15 British airports, with displaced aircraft and crews prolonging the trouble after flight processing resumed.

Industry estimates cited by British media put the number of disrupted passengers near 330,000, including more than 155,000 whose flights were canceled and more than 180,000 who were delayed. Heathrow, Gatwick, Manchester, Stansted and airports outside England all absorbed consequences, while some flights bound for Britain were held, diverted or canceled abroad. The geographic reach explains why a four-hour technical event can require several days of schedule repair.

The costs also extended beyond ticket changes. Airlines had to reposition aircraft, find legal duty time for crews, secure hotel rooms and arrange alternate routes during a period when seats were already committed. Travelers missed connections, work and family events. None of those effects determines technical responsibility, but they make resilience a consumer issue as much as an engineering one.

A System Tested Before

The comparison hanging over NATS is the August 2023 outage, when a flight plan containing two geographically separate waypoints with identical names caused the primary and backup systems to enter a fail-safe mode. NATS’s 2023 report said the system could not resolve the ambiguity automatically and that engineers took several hours to identify the unusual sequence.

An independent review subsequently produced 34 recommendations covering NATS, the Civil Aviation Authority, airlines, airports and the government. The government’s response said the recommendations addressed engineering resilience, crisis management, communications and passenger care. Regulators have said the recommendations were completed, making the latest failure a test of whether those reforms were broad enough rather than proof that they were ignored.

NATS says Tuesday’s event was different from both the 2023 breakdown and a shorter disruption in 2025. Airlines are skeptical. Ryanair Chief Executive Michael O’Leary said the carrier had initially been told that a rogue flight plan was involved and accused Rolfe of offering inconsistent accounts, according to Reuters coverage. His company is separately suing NATS for more than £7 million over losses from the 2023 outage, a financial dispute that should be weighed when considering its criticism.

The recurring question is not whether every rare input can be predicted. Safety-critical networks are expected to fail conservatively when they encounter data they cannot trust. The policy question is whether a conservative failure must also remove so much capacity, and for so long, that a national system loses thousands of flights before normal processing returns.

What Airlines Still Owe Passengers

The Civil Aviation Authority said disruption directly caused by the NATS outage, including its knock-on effects, would probably qualify as “extraordinary circumstances.” That classification makes standard cash compensation unlikely, although the regulator stressed in its passenger guidance that each case depends on its facts and travelers may still pursue claims.

Extraordinary circumstances do not erase airlines’ duty of care. For a canceled flight, carriers must offer a refund, rerouting at the earliest opportunity or rerouting later. During long delays, they must provide meals, refreshments and, when necessary, hotel accommodation. If an airline cannot arrange appropriate care or replacement travel, the authority expects it to reimburse reasonable costs, so passengers should preserve receipts and records of the options they were offered.

Those rules apply broadly to flights departing the United Kingdom and to certain flights arriving on British or European carriers. Travelers arriving on other foreign airlines may be covered differently. The practical lesson is to deal first with the operating airline, document expenses and avoid assuming that a lack of delay compensation also means a lack of refund, rerouting or care rights.

The Questions Investigators Must Answer

Transport Secretary Heidi Alexander gave NATS one week to submit a technical account and asked the Civil Aviation Authority to conduct an independent review. She told Parliament that safety must remain the overriding priority but said she did not believe the disruption was unavoidable. The authority has said its review will establish what happened and examine whether NATS is structured to provide a resilient service.

The inquiry should be able to answer several distinct questions. Investigators need to determine exactly what the aircraft’s flight plan contained; whether it complied with filing requirements; why automated validation did not isolate the data; what processing function failed; how quickly controllers shifted to contingency procedures; and what capacity was available while the principal system was unavailable. They also need to test the reported absence of a dedicated backup against the resilience commitments made after 2023.

Accountability will depend on that sequence. If the military submitted invalid data, procedures for state aircraft may need revision. If the filing was valid but unfamiliar, the central issue will be software validation and graceful degradation. If operational or management choices prolonged recovery, the review will have to distinguish those choices from the initial trigger. Assigning blame before those facts are established would obscure the engineering lessons travelers need authorities to learn.

For now, the evidence supports a narrower conclusion: a new, credible report has identified military flight data as the apparent trigger for Britain’s worst air-traffic disruption since 2023, while the government department associated with that aircraft disputes that it made an error. The coming reports must reconcile those positions and explain why one submission could remove enough processing capacity to cancel more than 2,000 flights. Until they do, the military link is a significant lead—not a settled verdict.