The short version: Chase attributed the November 16 disruption to a software bug and said it was not cyber-related. Zelle's operator said its network was operating, although transactions involving affected banks could be delayed. Check a pending payment's status before trying again, and contact your bank through a number you already trust.
The Chase Bank Outage Reveals a Bigger Problem With Digital Money
JPMorgan Chase says a software bug delayed some transactions on November 16, with effects reaching payments involving other banks. The incident exposes a question that matters well beyond one app: how much of modern banking depends on systems customers never see?

Jump to a section
A payment can leave one person's screen before it reaches another person's account. In between, several systems may need to agree on what happened. When one part slows down, the person waiting for the money does not see a diagram of those dependencies. They see a missing payment.
On November 16, customers reported trouble with online banking and payments involving several major US financial institutions. Chase said a software bug in an on-premises hardware system was delaying some transactions. A bank spokesperson told The Associated Press the issue was not cyber-related.
That is the bank's assessment, not a complete public technical investigation. As of 00:16 UTC on November 17, equivalent to 7:16 p.m. Eastern on November 16, the latest AP account we reviewed said Chase was working to resolve the issue and expected a fix by the end of Monday. Chase also said customers still had access to their accounts and funds. That statement and the anticipated fix do not establish that every customer could use every service or that every delayed payment had been reconciled.
One bank's delay can reach another bank
A person at one institution can send money to someone at another. If the destination or originating bank is having trouble processing its side, the sender may see a delay even when their own bank's application is working.
Chase told AP that other banks could experience slower transactions when they were party to a transaction with Chase. AP said customers at Bank of America, Citigroup and other banks could encounter such delays. That does not mean those banks experienced the same internal software bug.
There is an important distinction between a service being down and a payment through that service being delayed. Early Warning Services, which operates Zelle, told AP that its network was operational. A transfer involving a Chase customer could still be delayed if one of the participating institutions was having trouble.
Reports on outage-monitoring websites can help identify a problem, but they are reports from users, not a count of affected accounts or proof of a shared cause. Some customers also reported problems with other banks; the available statements do not establish that every report came from Chase's issue.
Was it a hack? What about AI?
Chase said the problem was a software bug and not cyber-related, according to AP. We have not found evidence that a cyberattack, an AI agent or AI-written code caused this banking incident. The bank has not published enough technical detail for an independent reader to reconstruct the failure.
The Microsoft 365 outage began only days earlier. Microsoft said that incident was a technical issue related to a network upgrade, not cyber activity or AI agents. AP reported that officials and company leaders said the disruptions were unrelated. Timing alone is not evidence of a common cause.
That does not make the comparison useless. It changes the question. Instead of asking whether the same attacker hit two systems, ask what happens when a service millions of people depend on becomes a weak link for everyone connected to it.
A separate UK outage exposed another dependency
There is a documented, separate example. On October 20, a power failure at a Google data-center facility disrupted cloud services in the United Kingdom. In a November 4 letter, the Bank of England said some financial firms experienced disruption to customer-facing banking and payment services as well as internal systems. The impact varied because firms used those services differently.
The regulator said most business-as-usual operations across the affected sector had resumed by the end of that day, although some customers experienced material disruption. That UK cloud incident and the Chase software bug have different reported causes. There is no established connection between them.
The useful connection is structural: a firm can have working parts of its own service and still depend on another institution, provider or infrastructure layer for a particular transaction. A resilient system needs to know which dependency has failed, what remains usable and how to reconcile unfinished work afterward.
The AI connection is a future risk, not a diagnosis
Banks are already using AI for tasks ranging from operations to security. That creates a separate question: if many institutions rely on the same small group of outside providers, could a problem at one provider affect several institutions at once?
The Bank of England raised that possibility in its July financial stability report. It described potential concentration risk if firms come to rely on a handful of advanced AI providers for cyber defense, and considered scenarios involving shared software or provider disruption. These are risk scenarios, not findings about the Chase outage.
In a 2026 Bank of England and Financial Conduct Authority survey, third-party implementations accounted for 39% of the AI systems reported by participating firms. The survey received 196 responses from UK financial firms. That figure is a share of reported systems, not a share of firms or US banks, and it does not establish what caused the Chase incident. Respondents also identified cyber risk and critical third-party dependencies as concerns as AI use expands.
AI could help teams find problems and defend systems. It could also add another service to the dependency map. The practical test is whether a bank can keep essential services safe, explain what is affected and recover transactions when a provider or internal component is unavailable. That is an operational question, not proof that AI failed in the November 16 incident.
If a payment is missing or delayed
A status such as pending, processing, failed or completed can mean different things across banks and payment methods. A blank or failed-looking screen does not necessarily show whether a request was submitted, and a completed status on the sender's side may not confirm when the recipient can use the money. Ask your bank what the status means for that payment and check with the recipient before sending it again.
- Check automatic payments. Ask your bank whether scheduled withdrawals or bill payments were submitted and when their status should become visible. Keep any confirmation you receive.
- Keep a private record. Note the amount, time, recipient, confirmation number and any error message. Keep screenshots securely; do not post account details or confirmation numbers on social media.
- Protect important deadlines. If rent, a bill or a loan payment is due, contact the recipient using a trusted number and ask what alternatives or extensions are available. Check with your bank before using another method if a payment may already be processing.
- Ask about fees. If a delay leads to an overdraft, late fee or other charge, document it and ask the bank or biller whether it can be reversed. A waiver is not guaranteed. The Consumer Financial Protection Bureau recommends checking automatic payments and asking about outage-related fees.
- Report unauthorized activity promptly. An outage is not proof of fraud, but an unfamiliar transfer deserves immediate attention. Contact your bank using the number on your card or statement. The CFPB explains how to report an unauthorized transaction; protections and deadlines depend on the transaction and circumstances.
If an app is unavailable, a branch, ATM, telephone service or another payment method may help, but availability can vary during an incident. Confirm what is working before relying on an alternative. If you have sent money to a scammer, our guide to reporting a scam payment to your bank explains what to document and ask.
Beware the fake recovery message
Outages create an opening for impersonators. A text or caller may claim you must verify your account, move money to a safe account, install software or provide a login code to restore access. Treat an unsolicited request like that as suspicious and contact your bank independently. The CFPB warns that messages asking for account information can be phishing.
If you shared credentials or see activity you do not recognize, contact the bank immediately through a trusted channel. Our bank account security and recovery guide covers further steps. Do not share your password or one-time login code with anyone offering to fix the outage.
When the app comes back
A working login is not the same as a reconciled payment. After service returns, check whether each pending transaction completed, failed or was duplicated, and follow up on anything that remains unclear.
The November 16 disruption is a reminder that digital banking is a chain of institutions and systems, not a single screen. The strongest test of that chain is not whether it never fails. It is whether people can understand the failure, avoid making it worse and see their transactions resolved when service returns.