FAA Outage: FBI Reportedly Examines Software Update
The FBI reportedly examined whether a software update caused the FAA outage that halted departures across the United States on November 13, 2026. That possibility remains an investigation lead, rather than an established cause. The reported inquiry also came with an important qualification: investigators did not initially believe a cyberattack or criminal activity caused the disruption.
Bloomberg attributed that assessment to a person familiar with the matter. SecurityWeek subsequently described the same reporting in its coverage of the software update inquiry. The distinction matters because a suspected maintenance problem, a confirmed software defect, and malicious interference describe different failures.
The central tension concerns routine equipment work meeting an aging network that must remain dependable throughout maintenance. The January 2023 nationwide grounding offers a reference point: ordinary technical work can have extraordinary operational consequences. For the Federal Aviation Administration and its network provider, L3Harris, the immediate challenge is explaining the failure without outrunning the evidence.
What Changed in the FAA Outage Investigation
The investigation moved from an unexplained communications disruption toward specific network activity, while the underlying cause remained unresolved.
The November 13 reporting identified a possible software update as an investigative lead. It did not establish which update investigators were examining. It also did not identify a confirmed defective software version, a completed forensic finding, or a criminal act.
By November 15, the FAA had published a more detailed operational account. Its explanation described excessive connection attempts destabilizing the FAA Telecommunications Infrastructure network, which carries communications supporting aviation operations. That account provides a narrower technical starting point than the broad phrase “computer outage.”
The agency identified a radar facility in Amarillo, Texas, and a radio communications facility in El Dorado, Arkansas, as potential origins. It also connected the inquiry to L3Harris equipment work. These locations are investigative leads, rather than a final allocation of responsibility.
In its ground stop explanation, the FAA said L3Harris had been deploying a new router at El Dorado. A router directs traffic between network connections. The agency said blocking the identified locations stabilized the network.
The FAA placed the initial technical problem at approximately 3:30 p.m. Eastern on November 13. Air traffic resumed at 8:15 p.m. that evening. Those timestamps describe the operational interruption, while the investigation continued afterward.
The agency also said this was not a complete loss of radio and radar capabilities. Some frequencies and radar feeds remained available. According to the FAA, established procedures allowed aircraft already airborne to land safely.
Taken together, those details support a specific account of disruption and recovery. They do not complete the chain from equipment change to nationwide instability. An intervention that restores service helps investigators narrow the problem, but it does not automatically identify every contributing condition.
That distinction prevents an early hypothesis from becoming an unsupported headline conclusion. A software update might be part of the eventual explanation. Hardware behavior, configuration choices, interactions with older equipment, or operating procedures remain questions requiring evidence.
The dates also matter. Information released on November 15 provides later context for reporting published on November 13. It should not be presented as something investigators had already established when the initial Bloomberg story appeared.
For readers following the FAA outage investigation, the most useful distinction is between observation and explanation. Investigators have operational observations about instability and recovery. A final explanation must connect those observations to a reproducible failure and an effective remedy.
Routine Maintenance Now Faces a Larger Reliability Test
The pressure falls on the FAA and L3Harris to show why a maintenance event reached beyond its immediate location.
The national consequences make this more than an equipment replacement story. A remote communications facility serves a local function within a connected system. When disruption extends beyond that facility, investigators must examine the boundaries intended to contain abnormal behavior.
The FAA described the work as routine equipment replacement. It also said it would reconsider when such work occurs. That acknowledgment raises an operational question without establishing negligence: were the existing maintenance assumptions appropriate for the network’s current condition?
Maintenance timing changes exposure, but timing alone cannot explain the failure. Moving work into quieter hours can reduce disruption to travelers. It does not necessarily prevent the same technical behavior from affecting connected facilities.
The deeper issue is whether the system can absorb a foreseeable equipment problem. A network supporting essential operations needs ways to detect, restrict, and recover from abnormal traffic. Whether those protections existed and performed as intended remains part of the unresolved engineering story.
For L3Harris, the required explanation reaches beyond whether technicians followed an approved procedure. Investigators also need to understand the procedure’s assumptions. A familiar task can become riskier when surrounding equipment, network dependencies, or operating conditions change.
For the FAA, oversight includes understanding those dependencies before maintenance begins. Contractor responsibility and agency responsibility can overlap without being interchangeable. Establishing either requires the actual work records, approval process, technical findings, and applicable contractual obligations.
There is also a difference between restoring service and restoring confidence. A successful intervention demonstrates that operators regained control of the immediate situation. Confidence requires evidence that the underlying failure cannot easily recur through another device or location.
Airlines face a separate recovery problem once departures resume. Aircraft, crews, passengers, and airport resources must return to workable schedules. Restored communications therefore do not mean every disrupted journey immediately returns to its original plan.
A traveler waiting at a departure gate experiences the consequences through changing departure estimates and missed connections. The investigation addresses a different question: which technical and operational protections failed before those consequences reached the airport?
Keeping those perspectives together makes accountability more concrete. Passengers need accurate recovery information, while engineers need a defensible causal explanation. Neither group benefits from a premature declaration that the problem was simply an update.
The immediate pressure is therefore practical. The FAA and L3Harris need to explain what happened, what they changed, and how they assessed remaining exposure. A broad promise of modernization cannot substitute for those incident-specific answers.
New Equipment Must Work With the Network Already Running
The central engineering conflict is between changing individual components and preserving dependable behavior across the existing system.
Replacing old equipment is necessary maintenance. It also creates an interface between new behavior and established dependencies. The relevant question is whether that interface was understood and tested under conditions resembling actual operations.
The FAA attributed the instability to its legacy TDM system. TDM, or time-division multiplexing, carries multiple communications streams through a shared connection in assigned time slots. That description identifies an older telecommunications environment, but it does not establish the precise defect.
It would be premature to translate that account into a specific software diagnosis. Excessive traffic can have different origins. Investigators need device records and network evidence before identifying whether software, configuration, equipment, or interacting conditions initiated the event.
A useful engineering reconstruction must separate the initiating event from the conditions that amplified it. The initiating event explains why abnormal behavior began. The amplifying conditions explain why that behavior reached services beyond its original location.
Those are related questions with different remedies. Fixing an initiating defect addresses a particular trigger. Improving containment addresses the possibility that another trigger produces similar consequences later.
This distinction applies even if investigators ultimately confirm a software update caused the initial problem. A confirmed update defect would explain an important part of the outage. It would still leave questions about testing, detection, isolation, and recovery.
Conversely, a finding that rules out a software defect would not eliminate the maintenance connection. Equipment installation can involve settings, physical connections, and interactions with existing systems. Those possibilities require examination rather than assumption.
The FAA has distinguished routine replacement work from its broader modernization program. That distinction is reasonable as a statement about program scope. It does not remove the need to understand how maintenance and modernization interact operationally.
A network cannot be evaluated solely through the age of each component. Dependencies matter because a newer device can exchange traffic with older equipment. Reliability depends on the behavior of that connected arrangement during normal operations and unexpected conditions.
The stronger technical question is therefore about compatibility under failure. Did testing examine what happened when a replacement device behaved unexpectedly? Did operators have enough visibility to identify the affected connections quickly?
These are analytical questions, rather than findings about what the FAA or L3Harris failed to do. Publicly available information does not answer them. They identify what a credible technical explanation should eventually resolve.
The network instability also illustrates why replacement progress and operational resilience are different measures. Installing equipment records completed work. Demonstrating that a failed component stays contained provides evidence about the dependability of the resulting system.
The FAA outage puts those measures beside each other. The agency must continue maintaining aging infrastructure while preserving current service. The task is to make changes without allowing a local problem to become a national interruption.
That challenge does not support stopping upgrades indefinitely. Older equipment creates its own maintenance and support risks. It supports examining the transition process with the same care applied to the replacement technology.
The 2023 Grounding Offers a Warning, Not a Diagnosis
The historical comparison concerns the reach of maintenance failures, rather than proof that the same technical problem happened again.
In January 2023, the FAA grounded departures after a failure involving its pilot-notice system. The Notice to Air Missions system, known as NOTAM, distributes information about conditions affecting flight operations. Its role differs from the telecommunications network involved in the November 2026 investigation.
The FAA’s January 19, 2023, statement said contract personnel had unintentionally deleted files while correcting synchronization between primary and backup databases. Its NOTAM outage findings also reported no evidence of malicious intent at that stage.
The comparison is useful because both incidents put ordinary technical work under extraordinary scrutiny. It is limited because the systems and reported failure paths differ. A database problem does not establish how a telecommunications disruption began.
There is a broader lesson in the sequence of the 2023 explanation. Early statements described a damaged database file. Later findings provided more detail about how files were deleted during maintenance.
That progression shows why initial descriptions should remain provisional. A visible symptom can be accurate without being a complete cause. Investigators often need additional records and reconstruction before connecting a symptom to the action that produced it.
Independent oversight also provides context beyond either incident. The Government Accountability Office, Congress’s audit agency, has documented longstanding problems sustaining FAA technology. Those findings establish background risk, rather than responsibility for the November disruption.
In September 2024, GAO reported that the FAA had classified 51 of 138 air traffic control systems as unsustainable. Another 54 were potentially unsustainable. Its aging systems assessment discussed problems including outdated functionality and unavailable spare parts.
Those classifications should not be treated as a count of systems that failed on November 13. They describe a wider portfolio assessment. Their relevance is that maintenance takes place inside an environment already facing documented support challenges.
The industry comparison here is between operating approaches rather than competing vendors. Continuing to maintain older systems preserves current service but retains existing limitations. Replacing them addresses those limitations while introducing transition work that must also be managed.
Neither approach offers an automatic escape from operational risk. Delayed replacement can increase dependence on difficult-to-support equipment. Accelerated replacement can create coordination problems if installation plans overlook shared dependencies.
GAO raised that second concern in its September 2026 review of modernization planning. The agency found shortcomings in cost estimation and scheduling. Those findings predated the November outage and should not be presented as a diagnosis of it.
The historical record therefore supports a narrower, defensible judgment. Aging infrastructure needs replacement, and replacement needs disciplined execution. The November investigation must establish which specific failures occurred before assigning the event to either category.
For developers and enterprise technology buyers, that judgment has a familiar application. A supported product can still depend on older infrastructure elsewhere. Evaluating a supplier’s reliability requires understanding those dependencies, rather than relying only on its newest features.
The Software Update Hypothesis Still Has an Evidence Gap
A reported investigative lead does not establish a confirmed cause, and an FBI inquiry does not establish criminal activity.
Bloomberg’s account relied on a person familiar with the matter. SecurityWeek attributed its description of the software update hypothesis to that reporting. That sourcing supports careful attribution, not an assertion that the FBI publicly confirmed a defective update.
The same account said the FBI did not believe a cyberattack or criminal activity caused the communications problem. That qualification belongs alongside the update hypothesis. Removing it would distort the reported state of the investigation.
The FAA separately said available evidence gave no indication of a cyberattack or malicious outside actor. That is an evidence-based assessment at a particular stage. It should not be expanded into a claim that every conceivable security question has been permanently closed.
Likewise, the involvement of federal investigators does not establish sabotage. Technical disruptions affecting essential infrastructure can require several kinds of expertise. The presence of law enforcement is not evidence of an attacker, a prosecution, or criminal intent.
There is also an attribution gap between equipment work and legal responsibility. The FAA identified L3Harris as its network provider and described work at a relevant location. Those facts do not independently establish breach of contract or misconduct.
A complete finding would need to connect the work to the failure through evidence. It would also need to consider the surrounding system and approved procedures. Responsibility cannot be inferred solely from whose equipment appeared in an early explanation.
The same caution applies to claims about artificial intelligence. The FAA said its SMART system was unrelated and did not feed information into air traffic control. There is no basis in the verified material to describe this as an AI-caused outage.
Nearby technology news can make unsupported connections attractive. A prominent launch or another service interruption supplies an easy narrative. Temporal proximity does not demonstrate a shared technical dependency or a causal link.
For the FAA outage, the unanswered questions remain concrete. What changed before instability began? Which behavior propagated, which protections responded, and which intervention addressed the initiating problem rather than its symptoms?
Public explanations do not need to expose sensitive operational details to answer those questions meaningfully. A summary can identify the failure category, contributing conditions, corrective actions, and validation process. That would allow outside assessment without publishing an operational blueprint.
The skeptical position is therefore straightforward: the current account explains more than the earliest alerts, but it remains incomplete. Restored operations are welcome evidence of recovery. They are not equivalent to a finished investigation.
That position also protects against an opposite overstatement. Uncertainty about the precise cause does not mean nothing is known. Investigators have narrowed the inquiry, and the FAA has described specific observations that future findings must explain.
A useful update should reduce that uncertainty with evidence. It should identify which earlier assumptions held, which changed, and why. Repeating the phrase “software update” without that detail would add familiarity without adding understanding.
Three Findings Will Matter More Than Another Broad Promise
The next useful developments are a documented cause, validated containment, and maintenance changes tied to the investigation’s findings.
First, watch for a technical finding that connects the original change to the observed network behavior. It should distinguish a software defect from configuration, equipment, compatibility, or procedural issues. Those categories imply different corrective actions.
Confirmation of the update hypothesis would strengthen the early reporting. A finding that identifies another cause would narrow or overturn that hypothesis. Either outcome would be more informative than continuing to describe the update as a possibility.
The strongest account would explain the sequence without confusing recovery actions with causes. What began the disruption, what allowed it to spread, and what stopped it? Those questions form a practical test for the completeness of the explanation.
Second, watch for evidence that containment measures work beyond the locations initially identified. Containment means limiting a failure’s effects so that other services remain dependable. An effective remedy should address comparable conditions elsewhere in the network.
A statement that equipment was adjusted establishes an action. Evidence that the adjustment prevents the relevant failure establishes something more useful. The distinction matters because an intervention can restore service without resolving every pathway to recurrence.
Validation should also explain what happens when abnormal behavior returns. A dependable system needs detection and recovery as well as prevention. Evidence of those capabilities would strengthen confidence in the response even before every replacement project finishes.
Third, watch for maintenance and modernization changes that reflect the actual findings. These should connect scheduling, testing, approval, and recovery decisions to the failure investigators establish. A changed maintenance window alone would leave the broader compatibility question open.
There is already an independent benchmark for assessing planning quality. GAO’s September 2026 modernization planning review found that the FAA lacked an integrated master schedule for its initial modernization phase. The report identified coordination risks before this incident occurred.
That review does not prove scheduling caused the November disruption. It does explain why coordinated implementation deserves attention alongside equipment purchases. Progress is more meaningful when installations account for operational dependencies and the effects of overlapping work.
Evidence of better coordination would strengthen the case that lessons have changed operations. Another general commitment to newer technology would leave that question unanswered. The useful measure is whether the response addresses the failure that investigators actually find.
For organizations following this story, the practical takeaway concerns evidence quality. Incident summaries should separate observations, working hypotheses, and established conclusions. That structure helps teams avoid turning an early explanation into a permanent assumption.
Keep following the FAA outage through those three signals: a documented cause, tested containment, and operational changes grounded in the findings. When a new update appears, compare its claims with the original timeline and the remaining gaps. Look for an explanation of how the failure began, why it spread, and how the remedy was checked. Apply the same standard to incident reports from your own technology suppliers. A report should help readers understand what changed and what evidence supports the response. Before accepting the next assurance that everything is fixed, ask which failure condition was reproduced and what demonstrated that it would remain contained.



