Key facts
- The president supported firing Sean Cairncross.
- Cairncross did not comment on leaving the job.
- Kirsten Davies, the Defense Department chief information officer, will serve as interim national cyber director.
- A Microsoft outage hit email services.
The White House dismissed national cyber director Sean Cairncross on Thursday. Technology-related outages have occurred across the United States. The disruptions included a nationwide halt to air travel and a Microsoft cloud outage on Nov. 12. A separate Microsoft outage hit e-mail services during the same period, while a PlayStation Network disruption struck on Nov. 18. The US National Security Agency’s network also experienced a temporary disruption.
The nation’s top intelligence official said on Wednesday that the outages were separate events. The official said they stemmed from maintenance and technical changes. The official also said they were not a coordinated or state-sponsored attack. The remarks came after days of disruption to transportation, communications, and intelligence systems across the country.
Cairncross managed federal cybersecurity policy and strategy and worked as a Republican National Committee executive. The president supported firing Sean Cairncross, and a person close to the White House said the White House fired him, speaking anonymously. Kirsten Davies, the Defense Department chief information officer, will serve as interim national cyber director, and the Senate, led by Democrats starting in January, must confirm the successor. Cairncross did not comment on leaving the job.
Axios was first to report Cairncross’s departure. The Guardian cited unnamed officials who said Stephen Miller wanted the change because Cairncross lacked the strong cyber credentials for the job, stating he “lacked the strong cyber credentials to perform the job.”
The Electronic Frontier Foundation published an analysis on Nov. 19, 2026 about clustered outages affecting e-mail, air travel, spending money, Apple devices, and signals intelligence. The foundation’s report came after a Microsoft cloud outage on Nov. 12 and a PlayStation Network disruption on Nov. 18. EFF says none of these outages have been directly linked to AI or hackers so far. According to the foundation, bank CEOs and the NSA briefed Congress and reportedly linked some outages to rushed patching of AI-discovered vulnerabilities.
Every software update carries risk, and businesses and public bodies now release more updates more quickly than ever before. Some updates concern critical infrastructure, so an outage there would be disruptive, yet unpatched vulnerabilities pose a greater danger. Vulnerabilities are software bugs that affect security. After discovery and disclosure of a vulnerability, the software maker typically releases a patch fixing it.
Security fixes in update notes are patches to vulnerabilities.
Infrastructure operators face a choice between fast patch deployment and slow, cautious patch deployment. Patching quickly helps keep systems secure. Yet every software update carries the risk that a change to a deployed system will have unintended consequences leading to a disastrous outage. Resilient networks, servers, and applications require investment in people, time, equipment, and processes such as maintenance, redundancy, staging, rollbacks, monitoring, and testing.
Decision-makers rarely find maintenance investments as attractive as new products, features, or layoffs. Deferred maintenance eventually produces immediate and dramatic costs. Technical debt refers to known issues left unaddressed that can multiply and interact to cause outages. It appears as manual routine steps, fragile programmer shortcuts, overdue operating system reboots, or unsupported networking hardware.

Well-resourced routine maintenance makes significant issues from major security updates rare.
The short-term need is a big investment in people and training to roll out patches promptly and safely. The long-term need is a big investment in software security to eliminate classes of vulnerabilities such as memory safety before developers write code. Organizations used to slow, cautious patching face a challenge in switching to quick patching. AI-enabled vulnerability disclosures mean everyone faces pressure to switch to quick patching at once.
Quick and cautious patching is possible if organizations invest in less technical debt and larger maintenance budgets. The measures are expensive and time-consuming but necessary. Investing in people and training for patching will pay dividends in reliability and security for governments responding to AI-enabled cybersecurity threats. The accelerated rollout of AI makes reinvestment in people and systems more critical for stability.
Apple is an EFF donor, and EFF donors do not influence the group’s agenda or positions. Users should weigh the trade-offs of depending on one provider for multiple services. Consumer cloud and device services bundle e-mail, messaging, file storage, and payments. That convenience creates a risk of losing access to them if the service goes offline or suspends an account.
Trying interoperable alternatives can reduce reliance on a single company.
The 3-2-1 backup rule means keeping three copies on two different kinds of storage, with one copy stored elsewhere. A personal device, cloud storage, and a local disk can meet the rule if the device keeps data on its own drives.
Organizations often schedule major system updates during the first three weeks of November ahead of Thanksgiving travel. November events include Black Friday, Cyber Monday, Giving Tuesday, and Surveillance Sunday. Scheduling updates early lets organizations prepare for holiday traffic and move outage risk to a period before those dates.
A possible spike in outages cannot easily be verified, and increased coverage can make outages seem more significant and interconnected than they are. After the Microsoft cloud outage on Nov. 12, smaller disruptions are more likely to be reported as part of a pattern.

Join the discussion