On the evening of November 16, 2026, online banking stopped working normally for customers of several of the largest banks in the United States. Search interest in the episode crossed one million queries in a single day, with related searches naming mobile applications, balance displays, wire transfers, and bill payment. For households the disruption was an inconvenience layered on top of anxiety about money. For businesses it was something sharper: payroll files waiting for release, supplier wires parked in queues, and card settlement reports that could not be reconciled. This article explains, under the law of the United States, which rules decide who bears the loss when a bank’s systems go dark. The method is general and applies beyond any single incident: identify the kind of account first, then the kind of transfer, then the written agreement, and only then the remedy. Readers should note what remains to be established about the November episode itself, including its technical cause, its exact duration at each institution, and whether any institution will offer compensation beyond what the law requires. Those facts belong to the banks and their supervisors to confirm.
I. An Outage Is a Banking Event Before It Is a Legal Event
A bank outage produces two kinds of harm at once. The first is operational: money cannot be seen or moved through the usual screens. The second is legal: deadlines keep running even while the screens are frozen. A payroll tax deposit, a time-sensitive wire to release goods, or a stop-payment instruction does not pause because an application shows an error message. Understanding the November 16 episode therefore requires separating what press reports described from what the law actually guarantees, and separating consumer accounts from business accounts, because the two live under different federal regimes.
A. What the November 16 Reports Described
Press coverage on November 16, 2026 described customers of several large American banks, including several household-name institutions, unable to reach accounts through online and mobile channels during the day. A national newspaper’s live reporting described outages touching four of the country’s largest banks on the same Monday. A wire-service report carried by a Boston television station attributed the delays to a technical issue at the country’s largest bank and was updated at 8:02 PM Eastern time that evening. A business magazine framed the episode as arriving days after separate technology incidents involving a major software company and the federal aviation system. One markets outlet reported that the country’s largest bank ruled out a cyberattack and attributed the failure to a software problem. These accounts remain press reports, not findings by a banking supervisor, and they should be read with that limit in mind. No federal banking agency had, as of this writing, published an official determination of the cause, and none of the institutions involved had published a full timeline with restoration times for each channel. What the reports establish is the shape of the event: a daytime loss of digital access at multiple household-name banks, on a business day, with mobile balance checks, transfers, and payment releases affected. What they do not establish is liability. An outage can result from failed change management, a vendor patch, capacity overload, or human error, and each of those paths leads to a different conversation about fault. Businesses reading these reports should therefore treat them as the trigger for preparation, not as proof of a claim. The useful response is to preserve evidence early: screenshots with timestamps, failed confirmation numbers, copies of queued payment files, correspondence with relationship managers, and records of substitute steps taken, such as sending a wire by telephone or paying a critical supplier by another rail. Courts deciding later disputes will care far more about this contemporaneous paper trail than about headlines.
B. Why the Legal Question Starts With the Kind of Account
American federal law does not give every bank customer the same toolkit. Consumer accounts sit under a protective federal statute, the Electronic Fund Transfer Act, and its implementing rule, Regulation E, which together impose short investigation deadlines, provisional credit, and capped liability. Business accounts sit largely outside that statute. When a company moves money by wire, the governing framework is Article 4A of the Uniform Commercial Code, as enacted state by state, which allocates loss through the concept of a commercially reasonable security procedure, plus the deposit and treasury management agreements the company signed. This split explains why two customers hit by the same outage can walk away with different rights. A household whose debit card transactions post twice during a system glitch can invoke error-resolution duties that force the bank to investigate within ten business days and provisionally return the money. A company whose half-million-dollar supplier wire never leaves the queue has no provisional-credit right under federal law and must look instead to its wire agreement, to Article 4A’s rules on execution and cancellation, and to whatever the bank promised about cut-off times and alternative channels. The distinction also shapes who must act fast. Under the consumer regime, the clock runs against the customer through statement-review deadlines measured in tens of days. Under the business regime, the decisive clock often runs in hours: wire agreements set daily cut-off times, fraud callbacks expire, and recall requests lose force once a beneficiary bank accepts and pays out. A treasurer who spends the outage refreshing a mobile application instead of calling the wire desk and switching to a backup channel may lose the only practical remedy available. The rest of this article follows that order. It sets out the consumer rules first, because they are the most generous and the most misunderstood, and then the business rules, because they decide the largest dollar losses, closing with three court decisions that show how judges actually assign responsibility when electronic banking fails.
II. Who Bears the Loss When Money Cannot Move
Loss allocation in American electronic banking follows written duties, not sympathy. Congress assigned consumers a defined set of bank duties with deadlines and damage multipliers. The states, through Article 4A, assigned businesses a bargain centered on security procedures: the bank must offer procedures reasonably fitted to the customer, and the customer that accepts them generally bears the fraud loss when the bank follows them in good faith. An outage scrambles both frameworks without suspending either. Transfers that should have executed sit unexecuted, instructions given through one channel must be remade through another, and records of what was ordered when become the evidence that matters most. The sections below state each duty as it appears in the texts in force, with the court decisions that give those texts their practical meaning.
A. Consumer Accounts: Short Deadlines and Provisional Credit
The Electronic Fund Transfer Act covers electronic transfers from consumer accounts, and its error-resolution section requires a bank that receives proper notice of an error to investigate, decide whether an error occurred, and report the result to the consumer within ten business days. The statute defines proper notice as identifying the consumer’s account, stating that it contains an error and its amount, and giving the reasons for the belief that an error occurred where applicable. When oral notice is given, the bank may require written confirmation within ten business days if it told the consumer of that requirement and the address at the time of the call. Instead of completing the investigation in ten days, the bank may provisionally return the disputed amount to the account within that period, allowing the consumer full use of the money while the investigation continues, and then conclude the investigation within forty-five days. If the bank finds no error, it must send an explanation within three business days after finishing and, on request, supply the documents it relied upon. A bank that skips provisional credit without a good-faith investigation or a reasonable basis for its position exposes itself to multiplied damages in a lawsuit. These duties sit in the section of the statute devoted to error resolution, and they are mirrored in Regulation E, the federal rule that restates them with operational detail, including the forty-five-day investigation window and the two-business-day notice of provisional credit. Consumer liability for unauthorized transfers is capped. The statute limits a consumer’s loss to fifty dollars where notice is timely, and it excuses later losses only where the bank proves those losses would not have occurred but for the consumer’s failure to report an earlier unauthorized transfer shown on a periodic statement within sixty days, extended for a reasonable time in extenuating circumstances such as long travel or hospitalization. Regulation E layers a second tier onto that structure: a consumer who reports the loss or theft of an access device within two business days faces no more than fifty dollars of liability, while later reporting can raise the ceiling to five hundred dollars where the bank proves the additional transfers would have been prevented by timely notice. The United States Court of Appeals for the Ninth Circuit gave these provisions their sharpest recent reading in a December 20, 2021 published decision, Widjaja v. JPMorgan Chase Bank, N.A., 21 F.4th 579, which reversed dismissal of a consumer’s statutory claim and sent it back for further proceedings. The court held that a consumer must report an unauthorized withdrawal within sixty days after the bank sends the monthly statement reflecting it, that notice reaching the bank from a third-party source does not excuse the consumer’s own failure to report, and that the consumer’s pleading survives where it plausibly suggests the later transfers would have happened even with timely notice. The practical application is that the consumer must report within sixty days, third-party notice does not substitute for the consumer’s own report, and the bank must prove later losses would not have occurred but for the failure to report. The limit of this 2021 consumer ruling is that it construes the sixty-day report and but-for causation rules for household accounts and does not extend error-resolution or provisional-credit rights to business accounts. The opinion therefore teaches two practical lessons for outage periods. First, statement review remains the consumer’s non-delegable habit: system chaos does not extend the sixty-day period, and a relative’s or a merchant’s call to the bank does not substitute for the customer’s own report. Second, banks carry a genuine burden of proof on causation for later transfers, which gives consumers leverage where the bank’s own records show the fraud would have continued regardless. A separate section of the same statute addresses the outage scenario most directly by defining when a bank is liable for failing to complete a transfer it was properly instructed to make. The bank is liable for failing to execute in the correct amount or on time unless an exception applies, such as insufficient funds, legal process freezing the account, or an exceeded credit limit. Critically, the statute provides two technical-failure defenses: the bank escapes liability where it proves by a preponderance of the evidence that the failure resulted from an act of God or circumstances beyond its control despite reasonable preventive care and required diligence, or from a technical malfunction already known to the consumer when the transfer was attempted. An institution that can show it maintained reasonable resilience measures and communicated the outage before customers attempted transfers stands on stronger ground than one that stayed silent while customers queued instructions into a dead channel. That structure is why preservation of outage notices, status-page captures, and timestamps matters for consumers as well as businesses.
B. Business Accounts: Security Procedures Decide Everything
A business account has no federal provisional-credit right and no fifty-dollar cap. Its protection comes from Article 4A of the Uniform Commercial Code, the uniform state-law framework for wholesale wire transfers that the Federal Reserve presents as the national text, and from the agreements the business signed when it enrolled in online banking and wire services. The core rule provides that a payment order received by the bank is effective against the customer, even if unauthorized, where the bank and customer agreed that authenticity would be tested through a security procedure, that procedure is a commercially reasonable method of guarding against unauthorized orders, and the bank proves it accepted the order in good faith and in compliance with the procedure and any customer instruction restricting acceptance. Commercial reasonableness is a question of law for the court, judged on the customer’s wishes expressed to the bank, the circumstances the bank knew including the size, type, and frequency of the customer’s usual orders, the alternative procedures offered, and the procedures generally used by similarly situated banks and customers. The commentary places the burden of offering adequate procedures on banks precisely because they design the systems and can best evaluate their effectiveness. Where the bank proves those elements, the business bears the fraud loss; where it cannot, the transfer is not effective against the customer, and a companion rule requires the bank to refund amounts collected on unauthorized or unenforceable orders with interest from the date it received payment. Two federal appellate decisions, three years apart, show both outcomes and together form the playbook for business outage and fraud disputes. In Patco Construction Co. v. People’s United Bank, 684 F.3d 197, decided July 3, 2012, the United States Court of Appeals, First Circuit, addressed a small Maine property-development business whose online banking credentials were compromised and from whose account seven days of fraudulent withdrawals in May 2009 totaled 588,851.26 dollars, of which the bank blocked or recovered part, leaving a residual loss of 345,444.43 dollars. The trial court had granted judgment to the bank, but the court of appeals reversed. It held that the bank’s collective failures, taken as a whole, made its security system commercially unreasonable: the system flagged every fraudulent transfer as high risk because the timing, value, and geography diverged from the customer’s profile, yet the bank monitored nothing, notified no one, and let the payments through, even though its own product offered device identification and risk scoring it never used. The court added that a generic one-size-fits-all configuration violates Article 4A’s instruction to account for the particular customer’s circumstances, particularly where lowering an authentication threshold to one dollar multiplied challenge questions for a customer making regular large transfers without adding monitoring in return. The practical application is that a bank must monitor the risk scores its own system produces and warn the customer when its own profile flags a transfer as high risk instead of letting the payment through silently. The limit of this 2012 ruling is that it turns on a bank that ignored its own high-risk flags under 2009-era authentication and does not decide liability for a pure outage with no unauthorized order. The case was remanded, with related claims reinstated.
In Choice Escrow & Land Title, LLC v. BancorpSouth Bank, 754 F.3d 611, decided June 11, 2014, the United States Court of Appeals, Eighth Circuit, reached the opposite result on different facts. An unknown intruder entered a title company’s online account after an employee fell for a phishing message, and on March 17, 2010 ordered a 440,000-dollar wire to an account in Cyprus, which the bank executed. The court affirmed judgment for the bank, emphasizing that the customer had declined stronger authentication the bank offered and that the bank accepted the order in good faith under the agreed procedure. The practical application is that a customer that declines stronger authentication the bank offered bears the fraud loss where the bank accepts the order in good faith under the agreed procedure. The limit of this 2014 ruling is that it rewards a bank that offered stronger tokens the customer declined and does not protect a bank whose own procedure was unreasonable or ignored. Read together, the two rulings tell a business exactly what to do before and during an outage. Before: accept the strongest authentication offered, enroll every authorized user in dual approval and callback verification, and keep the bank informed of normal transfer patterns so anomaly detection has a baseline. During: treat loss of digital access as a trigger to switch channels immediately, confirm every queued instruction by telephone with the wire desk, and document each step, because Article 4A rewards the customer who can show what was ordered, when, and through which procedure. Supervisory guidance points the same way from the bank’s side. A 2020 interagency bulletin from the Office of the Comptroller of the Currency, issued jointly with the Federal Reserve Board and the deposit insurer, sets sound practices for operational resilience at the largest banking organizations, calling for governance, scenario analysis, and secure, resilient operations. While directed at institutions above high asset thresholds, the paper states standards drawn from existing rules and guidance that examiners already apply, and it confirms that resilience planning, including disruption scenarios, belongs inside the bank’s duties rather than outside them. For a business negotiating treasury services, that guidance is leverage: ask how outage scenarios are tested, what manual fallback the agreement guarantees, and how quickly telephone or branch channels replace a failed portal. For broader context on the consumer rulebook that shadows these business disputes, the federal consumer agency maintains the official Regulation E hub that collects the rule, its history, and its interpretations in one place. Insurance deserves the same advance attention as banking channels. Many commercial crime and cyber policies respond to fraudulent-transfer losses only where the insured followed defined verification steps, such as telephone callbacks to known numbers before releasing wires above a threshold. An outage that forces staff onto unfamiliar manual procedures is precisely when those conditions are hardest to meet, so the policy wording should be tested against the fallback process before it is needed, not after a loss. Contract negotiation closes the loop: treasury agreements often set liability caps, notice windows measured in days, and cut-off times that assume a working portal, and each of those terms can be renegotiated at renewal to add outage-specific fallbacks, alternate-channel confirmation rights, and fee reversals for bank-caused delays. Practical preparation therefore converges on a short checklist that works under either regime: maintain a second payment rail with tested credentials, keep the bank’s wire-desk and fraud telephone numbers outside the portal, pre-authorize dual control and dollar thresholds in writing, photograph and log every failure contemporaneously, give the bank prompt written notice of any error or suspected unauthorized instruction, and review the account statements the moment access returns, because the sixty-day consumer clock and the business duty to report discrepancies both start from those records. None of these steps requires predicting the next outage. Each of them converts a screen that says nothing into a record that proves something, and in electronic banking disputes the record is usually the case. Businesses that want a Paris-based external perspective on structuring these cross-border banking relationships can turn to the firm’s business law practice in Paris for general information on how French companies organize their American banking documentation.
Conclusion
A bank outage tests preparation more than doctrine. For consumer accounts, federal law answers with investigation deadlines, provisional credit, capped liability, and a technical-malfunction defense that rewards institutions able to prove reasonable care and timely communication. For business accounts, the answer runs through security procedures: courts place the loss on the bank whose system ignored its own warnings and on the customer who declined stronger protection the bank offered. The November 2026 episode illustrates the pattern without settling any individual claim, since its cause, duration, and institutional responses remain to be officially established. The durable lesson is procedural and within every company’s control: keep a backup payment channel ready, confirm instructions outside the failed portal, write everything down while it happens, notify the bank promptly, and read the statements when access returns. Applied consistently, that discipline turns the next evening of frozen screens from a scramble into a documented file, which is the form in which electronic banking disputes are won or lost.
Read also
- DVLA Says You Are Medically Unfit to Drive at 70+: How British Residents in France Stay Licensed, Get a French Licence and Challenge the Decision
- British Husband or Wife of a French Citizen After Brexit: Getting the Vie Privee et Familiale Card, Moving to the Ten-Year Resident Card, and Challenging a Refusal or Removal Order
- British Pensioner With a French Pension Too: Why Your S1 Does Not Spare You French CSG-CRDS, and the Narrow Cases Where You Can Still Recover