The Follow-Up, Filed From the Record
The November 12–18 Technology Disruptions: What Failed, What Recovered, and What the Record Shows About the FAA, United, Microsoft, Apple, the NSA, and the Banks
A follow-up on the nationwide departure stop, the separate United Airlines outage, service disruptions at Microsoft and Apple, Chase transaction delays and reports at other banks, the classified NSA network interruption, and the Optimizely compromise. The record gives different answers about recovery, cause and scope.
The Record in Brief
Microsoft and Apple said their affected services had recovered. The Federal Aviation Administration says it resumed air traffic Friday evening and continued investigating the initiating fault. United said it resolved a separate interruption Monday. Chase acknowledged a software problem that delayed some transactions involving other banks, while several of those institutions reported their own systems were working. The National Security Agency acknowledged degraded internal service and said mission-critical capabilities continued. Its public statement gives no full-restoration time. A separate incident at Optimizely involved unauthorized code, which the provider says it removed.
Thursday through Saturday, November 12–14 | Microsoft 365
On Thursday, Microsoft customers reported problems with Outlook email, Teams and Copilot. The Associated Press covered the complaints as they were coming in. Reuters reported that the disruption affected communications beyond the United States, with Workhorse Group rescheduling its results release and Switzerland’s Richemont briefly delaying an announcement. The latter two are named consequences. App complaint counts are a different measure from the number of organizations or people affected.
Microsoft attributed the incident to a technical problem. Reuters reported that Microsoft knew of no malicious activity and ruled out AI as the cause. On Saturday Microsoft said it had addressed the underlying issue and restored the remaining affected customers and regions. The reporting establishes a multiday incident across products and regions, with recovery announced Saturday. The public accounts cited here do not show a continuous outage of every named product for every customer. Associated Press, initial Microsoft outage Reuters, Microsoft restoration
Friday, November 13 | FAA nationwide departure stop
The FAA dates the onset of its telecommunications trouble to about 3:30 p.m. Eastern. The order followed at 3:54 p.m., according to the FAA’s own advisory chronology. At 8:15 p.m. the agency says traffic resumed. The four hours and twenty-one minutes measure the order, from issuance to lifting. The measure from the first network trouble to resumed traffic is longer. The order held departures. The FAA says about 3,500 aircraft landed safely and certain radio frequencies and radar feeds remained usable, including for critical flights.
The FAA says instability reached 17 of its 22 en route centers and spread to several approach-control facilities. Its preliminary account identified two possible points associated with the excess network traffic: a radar site in Amarillo, Texas, and a radio communications site in El Dorado, Arkansas. Its network provider, L3Harris, had been installing a router at El Dorado. The agency says that isolating both sites stabilized the network and that it disabled a function on other routers and paused installations. Those actions establish what operators did to restore service. The agency’s Sunday account still described the initiating cause as under investigation and said the evidence then available gave it no indication of a cyberattack.
The FAA counted roughly 12,462 cancellations Friday, of which 9,082 were by major airlines, and expressly said that some cancellations in that total were not directly attributable to the stop. The traffic restart and the airlines’ return to their planned schedules were separate stages of recovery. The other counters printed their own totals for the same day, and they stand unblended: FlightAware’s later full-day count was 9,147 cancellations, printed in a later update as about 9,159, and Cirium counted more than 8,200 on Friday evening. Federal Aviation Administration, November 15 FAQ
Monday, November 16 | United Airlines
United’s departing flights were delayed by a separate interruption that lasted more than 90 minutes, according to the Associated Press. The airline reported that the outage was resolved Monday. The mechanism, as an Air Line Pilots Association memo seen by Reuters and reported by ABC News describes it, was a companywide outage that hit pilot applications, including crew communications, and cut off the airline’s ability to provide final aircraft weights for departing flights. Counts of Monday’s delays are snapshots taken at different times and can include delays with other causes. The AP also describes a United interruption in August 2025 tied to a system called Unimatic. That earlier incident’s delay total and system diagnosis belong to August 2025. Associated Press, United Airlines
Monday into Tuesday | Chase, Wells Fargo, Citi and Zelle
Chase said a software fault in its systems slowed some transactions. The bank said the effects extended to other institutions when Chase was a party to the transaction and said it saw a technical problem rather than a cyber incident. Customers described trouble viewing balances, using apps and completing transfers. CBS reported that complaint levels for the banks and Zelle had returned to normal early Tuesday. That measure describes reports received by the tracker. The cited statements do not give a final accounting of every delayed transfer.
Wells Fargo customers reported sign-in, balance and transaction problems. CBS sought Wells Fargo’s comment. The coverage reviewed for this report does not establish a Wells Fargo internal failure or a separate restoration time. For Citi or Citibank the reviewed public record likewise supplies no institution-specific confirmation of an outage, cause or recovery time. Those are two different evidence positions: customer complaints reported at Wells Fargo and an unconfirmed institution-level claim for Citi.
Bank of America said its systems operated normally. Capital One also said it was unaffected. U.S. Bank said its systems were not directly affected but warned that clients sending or receiving Zelle payments with another institution could face delays. Truist was among the names generating complaints in the outage tracker. Zelle said its service was operating and that issues at participating institutions could delay payments. A delayed transfer can leave a customer at either end reporting a service problem even if the network and the other bank remain available. The institutions’ statements and the users’ accounts are filed together below. CNN via ABC17, Monday’s bank reports CBS News, bank and payment complaints
Tuesday, November 17 | NSANet
The NSA acknowledged temporary degradation during scheduled maintenance. CNN identified the affected network as NSANet, used to share classified information internally, and Reuters carried the agency’s statement. The agency said some internal systems were affected and mission-critical capabilities remained operational. The Wall Street Journal separately reported that a person familiar with the incident described software patching and said there was no evidence of an attack. Reuters could not independently establish the scope or duration. The public accounts reviewed here give no confirmed restoration hour for the affected internal network. Continuity of the capabilities covered by the statement establishes a narrower point than full recovery. Reuters, NSA maintenance statement The Wall Street Journal, classified network reporting
Tuesday | Apple services, Schwab and Marriott
Apple attributed the disruption to an internal update and said it was not a cyberattack. The AP reported status-page trouble with Apple Music, the App Store, AppleCare and iCloud sign-in. Reuters reported problems with iCloud and Apple Pay while the company was still restoring services. CBS subsequently reported that Apple said its status page showed full recovery at about 5:46 p.m. Eastern. The earlier statement referred to partial recovery. No comparable count of all affected customers was supplied. Associated Press, Apple and Marriott Reuters, Apple during restoration CBS News, Apple restoration
Charles Schwab resolved a brief webpage issue, according to Barron’s. Business Insider described an error shown on the site and corrected the date to Tuesday. The material reviewed does not quantify how many clients were unable to trade or demonstrate a broader brokerage-system failure. Marriott told the AP it resolved a technical issue affecting sign-in on Marriott.com and in the Marriott Bonvoy app. The company statement addresses sign-in and gives no systemwide hotel check-in count. Barron’s, Schwab web issue Business Insider, Schwab website Associated Press, Apple and Marriott
The desk found no operator account in the material reviewed establishing that Workday, Cloudflare or National Weather Service reports were part of the incidents described above. Their inclusion in outage lists supplies neither a common cause nor a verified restoration time.
Monday and Tuesday | Optimizely’s confirmed security incident
Optimizely says someone added malicious JavaScript to some customers’ Web Experimentation projects without authorization. The provider says it fixed the vulnerability, removed injected code and took steps to secure affected accounts. Scott Helme of Report URI reported that the code was served through the platform on sites belonging to 178 organizations, naming Zoom, HSBC, Equifax, the BBC, Salesforce and Air France-KLM among them. He reported evidence of cryptocurrency theft and described risks to payment information and visitors’ devices. The figure counts organizations whose sites the researcher identified. It is not a count of compromised corporate networks, individual victims or bank accounts. The provider’s removal statement describes containment on its service. The ultimate scope of downstream loss remains open in the cited record. Optimizely, incident history Report URI, Scott Helme’s Optimizely investigation
Friday advisory | The FBI’s ARTEX warning
The FBI’s November 13 FLASH bulletin describes malicious use of an AI-enabled security tool called ARTEX in campaigns dating back at least to August. Its count of more than 1,600 publicly exposed instances, attributed to researchers, describes observed installations of the tool. It supplies no equivalent count of successful attacks. The bulletin is evidence that the FBI issued a warning on Friday about activity over a longer period. It supplies no link to the FAA interruption, the banking incidents or the named platform outages. FBI, November 13 FLASH advisory
Wednesday, November 18 | The status of the record
The Wednesday follow-up in The Wall Street Journal describes different changes at the organizations that experienced the outages. Granada Hoy’s Spanish-language roundup attributes the absence of public evidence of a shared attack to officials speaking in Reuters’ reporting. That republication is useful as coverage of how the story traveled, but it does not supply a second official investigation. La Tercera reports demands for explanations from lawmakers. Requests for explanations do not establish that investigators found a link or that a briefing took place.
The public record supports different recovery dates and different degrees of certainty. Microsoft, Apple, United, Marriott and Schwab reported restoration for the services described. The FAA reported stable operations and an ongoing root-cause investigation. In banking, the strongest cross-institution evidence concerns transactions involving Chase, with complaint levels falling by Tuesday. The NSA gave an assurance about mission-critical capabilities and left the duration of the internal outage unspecified. Optimizely separately confirmed malicious activity and reported removing the code. The reviewed record establishes no common actor, common dependency or common initiating event across those files. The Wall Street Journal, November 18 follow-up Granada Hoy, Spanish-language roundup La Tercera, Spanish-language U.S. outage coverage