Apple, the NSA, Chase, Microsoft, the FAA and United Airlines have all suffered disruptions within days. After this summer’s hacks of water systems and warnings about energy infrastructure, are outages becoming the new normal, or a preview of something far more catastrophic?

By Omar Afra

Getting your email, catching a flight, moving your money and signing into your cloud account have all become less dependable over the past week. A classified network at the National Security Agency also suffered disruption. For a public watching that sequence unfold, wondering whether America is under cyberattack is a reasonable response. (Reuters; NBC News)

The public record does not establish a coordinated attack behind this week’s failures. The companies and agencies involved have described software defects, maintenance problems and network instability, with several explicitly rejecting a cyberattack explanation. Even taken at face value, those accounts raise an uncomfortable question: why can routine work disrupt so much? (Microsoft incident messages; FAA)

If an ordinary update can knock essential services offline, knowing that no adversary caused the failure offers limited reassurance. We still need to know how the failure spread, why the safeguards did not contain it, and what would happen if someone deliberately tried to prevent recovery.

How Much Has to Break Before We Stop Calling It a Glitch?

The dates matter. This summer’s water attacks and this week’s technology failures belong in the same discussion about vulnerability, but their causes must be examined separately.

Reported incidents and explanations as of November 17, 2026
Date in 2026SystemWhat is on the record
July 26 and 27Minnesota water utilitiesMalicious activity targeted more than 30 community systems. Being affected did not necessarily mean a community lost water service. (MNIT; MNIT update)
November 12 to 13Microsoft 365Exchange Online and dependent functions suffered access failures. Microsoft identified a software defect exposed during network maintenance. (Microsoft)
November 13FAANetwork instability prompted a nationwide departure halt. The FAA reported no indication of an attack; its investigation into the underlying cause continued. (FAA)
November 16ChaseA software problem disrupted transactions, including some involving other banks. Chase said it was not caused by a cyberattack. Zelle said its own platform remained operational. (NBC News)
November 16United AirlinesA system outage delayed departures for more than 90 minutes. United reported no indication of a cybersecurity incident. (Associated Press)
November 17AppleServices including iCloud and Apple Pay were affected. Apple attributed the disruption to an internal update and said it was not a cyberattack. (Reuters)
November 17NSAA classified internal network suffered disruption. The agency said maintenance affected some systems while essential national security capabilities continued operating. (NBC News; NSA)

A cluster of outages can look like a pattern before anyone has established a common cause. It can also expose weaknesses that deserve attention regardless of who, if anyone, exploited them. The test is whether the explanations survive examination of the incident timelines, system dependencies and recovery records.

The U.S. map within cyan computer code flickers with abrupt blackouts, missing bands and broken signal effects.
AI-generated illustration for Testset.

Who Says Iran Did It, and What Can They Prove?

The summer water attacks make the stakes physical. In its July 30 warning about incidents across at least seven states, the FBI described intruders altering settings on industrial controllers and disrupting operators’ ability to monitor or control equipment. Reported consequences included lost water pressure and flooding. (FBI and EPA)

The New York Times reported on July 30 that investigators considered Iranian hackers the likely perpetrators of the Minnesota attacks. Its report explicitly called that assessment preliminary and said it could change as more technical evidence emerged. That qualification belongs alongside the allegation. (The New York Times, syndicated by GV Wire)

There is subsequent evidence to consider, too. In October 1 testimony to the New York State Senate, CSIS researcher Lauryn Williams said the CyberAv3ngers group, which has been linked to Iran, had claimed responsibility, while noting that the intelligence community had not yet formally attributed the summer attacks. A claim of responsibility is relevant; investigators still have to corroborate it. (Written testimony)

Those are serious grounds for investigating an Iranian connection. They also leave legitimate questions: what technical evidence supports the attribution, what can be independently assessed, and did the same actor carry out every incident? Minnesota’s own July 30 statement said investigators had not established that all the incidents shared an attacker. (MNIT)

Attributing attacks on civilian infrastructure can carry consequences far beyond a computer network. Officials should explain their confidence and the evidence they can responsibly disclose. An intelligence assessment deserves examination even when its geopolitical explanation sounds plausible.

Who’s Holding the Israeli Contractor Accountable?

An Israeli company’s role in Minnesota water cybersecurity is documented. Waterfall Security Solutions supplies security technology to Saint Paul Regional Water Services. The documented agreement covers a security gateway and support for St. Paul’s utility. It does not establish that the company operated the water systems hit across Minnesota. (MPR News; City contract record)

There is also a documented dispute about retaining the company. In September 2025, MPR reported that the water agency’s general manager credited Waterfall’s device with protecting its data transfers during a separate cyberattack on St. Paul earlier that year. That performance claim is a proper subject for scrutiny. What independent evidence supports it, and what exactly did the equipment protect? (MPR News)

The later contract decision has a clear date. On July 14, 2026, the water board approved an amendment setting a term of five years through August 31, 2031, with a stated maximum amendment cost of $70,376.25. The resolution says Waterfall prevailed in a competitive proposal evaluation. That approval came 12 days before the water attacks on July 26 and 27. Those later attacks cannot explain an approval already given. (City resolution; Attack dates)

The earlier attack, the officials’ claims about Waterfall’s performance, and the later contract extension provide plenty to examine without collapsing separate events into one story. Ratepayers should be able to inspect the procurement criteria, competing evaluations and evidence supporting the security claims. They should know what support access a supplier has, what its contract excludes, and how difficult replacing it would be.

Geopolitical targeting is a real issue for procurement to consider. In 2023, federal agencies documented hackers linked to Iran targeting Unitronics controllers made in Israel in American water facilities. Unitronics and Waterfall are different companies. The FBI’s July 2026 water advisory specifically identified Rockwell Automation/Allen-Bradley controllers. The public records reviewed here do not connect Waterfall to those intrusions. (2023 federal advisory; 2026 FBI advisory)

That leaves a substantial accountability question: how do public agencies evaluate a supplier’s security performance, jurisdiction, support dependencies and exposure to retaliatory targeting? Those assessments should be available for scrutiny whether the supplier is Israeli, American or based anywhere else.

What Happens When the Systems Don’t Come Back?

Energy infrastructure belongs in this discussion. A July 22 federal warning described activity by actors affiliated with Iran against industrial controllers across sectors including water and energy, causing operational disruption and financial losses. It did not establish a nationwide electricity blackout or connect those incidents to November’s failures. (CISA and partner agencies)

The difference between an interruption and a disaster depends on duration, damage and the ability to keep essential functions running. A prolonged loss of power, communications or control systems could turn a manageable incident into overlapping emergencies. That is a scenario to prepare for, not a forecast established by this week’s headlines.

The questions are practical. How long can a utility operate safely by hand? Have backup systems been tested under realistic conditions? Does a recovery plan depend on the same equipment or provider that failed? The FBI’s warning to water utilities specifically calls for practicing manual operations and routinely testing recovery arrangements. (FBI and EPA)

The FAA, the banks and the technology companies should face comparable questions about the services they provide. How did the failure escape its intended boundaries? Which safeguards worked? Which did not? What changed afterward, who checked the changes, and what happens if the same failure returns?

A bad week alone cannot establish that outages are becoming more frequent. Answering that requires comparable records of incidents, duration and people affected. But normalizing disruption before demanding those records would be a choice, and a bad one.

America already has documented attacks on essential infrastructure. This week’s outages add urgent questions about reliability, oversight and recovery. When services come back, the public deserves incident reports, independently assessed corrective work and clear responsibility for failures. A green light on a status dashboard tells us something is working again. It does not tell us how long we could function if it stopped.