Monday · Nov 16, 2026 · Eastern Time · Research cut-off ~4:40 PM ET · Updates to Tuesday 9:38 AM ET
What actually happened to every bank?
At 4:13 PM Downdetector's “Most reported right now” board listed 48 services — Chase at #1, Bank of America at #2, and 28+ financial brands in total. It looked like everybody went down at once. The official record says something narrower, and more specific.
Confirmed core: ChasePeak: ~1–2 PM ET, not 4 PMSource type: Official statements + wires citing DowndetectorCaution: Downdetector counts user reports — not affected customers, and not proof a service was down
The verdict
There was a real bank disruption today — but it did not start at 4 PM, it is not proven to be one internet-wide event, and the best-supported core is a Chase-centred technical bug that rippled through Zelle.
Chase confirmed a technical issue / “bug,” explicitly not cyber-related. Wells Fargo confirmed its own online-banking problem, with no cause stated. U.S. Bank said it was not directly impacted. Zelle said its network was fully operational. The Associated Press (updated ~3:58 PM ET) said Bank of America and Citigroup were not having issues with their own banking services. No shared cloud, DNS, BGP, card-network or Federal Reserve failure was found on any official timeline for Nov 16.
Updates
One flow, newest first. Start with the gauge — how far above normal the reporting actually is — then the board updates that led here.
Tue 9:38 AM · Chase Is ThirdTue 9:25 AM · Suspect № 1Tue 8:30 AM · The resolution7:45 PM · Second reading6:45 PM · The gauge5:20 PM4:55 PM
Tue 9:38 AM ET · Newest update
Chase Is Fixed. Chase Is Third.
Update — Tuesday, 9:38 AM ET · Latest
The 9:30 board turned over its top five — and kept the bank whose event ended yesterday evening.
At 9:30 this morning the board still holds 48 entries, and its top five has a new composition: Microsoft Outlook, Microsoft 365, Chase, My Toyota — and Grok, in the top five for the first time in this watch. Bank of America has left the five. Lower down, the rotation continues: Minecraft, Bluesky and Google Gemini are new to the board; Credit One Bank, Spotify and Yahoo Mail have re-entered; CNN — one of the 178 organisations on yesterday’s Optimizely list — has dropped off. Amazon Web Services is back on the board at № 38, showing “no problems.”
The top five — 9:30 AM
Downdetector most-reported board · Tuesday · 48 entries in total
1Microsoft OutlookOffice suite
2Microsoft 365Office suite
3ChaseBankResolved Mon ~6:30 PM
4My ToyotaCar app
5GrokChatbotFirst time in the five
Two office suites, one resolved bank, one car app, one chatbot. The board at 9:30 AM ET, Tuesday.
Read it against what is confirmed. Chase’s transactions have been posting normally since about 6:30 PM yesterday. The last infected Optimizely snippet was cleaned by 4:52 PM yesterday. Both of Monday’s fires are out. And the seismograph still reads weather: Chase third, a day later, on a board that has not dropped below 48 entries at any hourly check since Monday afternoon. What the board measures now is the residue — genuine friction at services with no announced problems, the aftershock of two public failures, and a reporting public that spent Monday learning to file a report first and ask questions later. The instrument is honest about what it is: user reports, not confirmed outages. It simply no longer describes a single event, because there isn’t one anymore.
There is a mood with a leaderboard.
Tue 9:25 AM ET · Suspect № 1
Suspect № 1
Update — Tuesday, 9:25 AM ET
The board never cleared because Monday had two events. The second one has a name now.
The second event
This morning Report URI published its teardown: on Monday, November 16, malicious JavaScript was served to visitors of 178 organisations through the Optimizely Web experimentation platform — Zoom, the BBC, CNN, Salesforce, GitLab, IKEA, HSBC, Equifax, Air France-KLM, Ledger and Peacock among them. The attacker wrote their code into the Project JavaScript field of customer projects; the rest of each snippet was untouched, still served from cdn.optimizely.com, under Optimizely’s certificate, from a host every affected site already trusted. How the attacker obtained write access to so many unrelated projects is unknown. Optimizely’s own sites carried the code too.
The kit
Every visitor got the same script, and all of it was live at once. A crypto clipper swapped wallet addresses — Bitcoin, Ethereum, Tron, Solana and more — in page text, form fields, the clipboard and outgoing requests, checksum-validated so it only replaced real addresses. A card skimmer, added at 13:43 UTC, watched for card numbers in sites’ own form fields (hosted payment iframes such as Stripe’s were out of its reach). On pages mentioning wallets or payments it exfiltrated cookies and a screenshot. And on iPhones and iPads it loaded two exploit kits — Coruna for iOS 13 to 17, DarkSword for 18.x — in a hidden iframe; pages on iOS 18.6 or earlier reloaded every 60 seconds to retry the exploit, a visible symptom any visitor could see. About $7,040 in cryptocurrency had verifiably reached the attacker’s wallets within nine hours, swept to exchanges at 13:03 UTC. How many cards and cookies were taken cannot be seen. The payload was rebuilt 26 times during the day.
One day, two fires
Monday, Eastern Time · two bands, one clock. Numbered markers are keyed below.
Infection — Optimizely snippets11:21 PM Sun → 4:52 PM Mon
123
Chase — reports & resolutionPeak 4,712 at 2:04 PM
4
Sun 11 PMMon 4 AM8 AM12 PM4 PM6:30 PM
18:03 AM — stolen crypto swept to exchanges (13:03 UTC).
28:43 AM — card skimmer added (13:43 UTC).
33:40–4:10 PM — mass cleanup of the snippets; last infected snippet found 4:52 PM (21:52 UTC). First injection was 11:21 PM Sunday (04:21 UTC).
42:04 PM — Chase reports peak at 4,712, after climbing through the morning; transactions posting normally by ~6:30 PM.
The malware was live before the first Chase report, through the peak, and past the cleanup. The two events share a calendar, not a cause — officials attribute the Chase outage to a software release, and nothing in the teardown touches that finding.
Why it is suspect № 1
This page’s open question was never Chase. Chase explained itself. The open question was the board: 48 entries at every hourly check through the night, brands rotating with no outage of their own, the airlines clearing by 1:30 AM while banks re-consolidated at 3, 5, 7 and 8, CNN appearing at № 48 by Tuesday morning. Now read the infected list against the board’s population:
Banking & finance32
Technology23
Retail20
Streaming12
News & media10
32 banking and finance organisations (HSBC in seven countries, Equifax, Paychex), 23 technology, 20 retail, 12 streaming, 10 news and media. A kit that reloads iPhone pages every sixty seconds and silently breaks copy-and-paste on payment pages manufactures “this site isn’t working” reports at industrial scale, on infrastructure that was never down. That is the leading explanation for the board’s breadth and its refusal to clear — stated as what it is: an inference from the overlap of the two lists and the shared clock, not a confirmed attribution. The board, it turns out, was a seismograph with two earthquakes on it.
178.
178 — organisations whose pages served the kit on Monday. 1,611 snippets checked; 188 infected.
The file
None of the 178 can produce a record of what their own pages served on Monday, because the script was a vendor’s and mutable by design — that is what a snippet is. The evidence in Report URI’s teardown exists because they fetched 1,611 snippets and kept a hashed copy of every infected one, with its headers and fetch time. Detection did not come from the brands, the vendor, or the platforms. It came from an outside watcher keeping sealed copies of a thing everyone else merely trusted.
Trusted JavaScript does not remain trustworthy. Somebody has to keep the file.
Tue 8:30 AM ET · The resolution
Update — Tuesday, 8:30 AM ET
The resolution, and the board that didn’t clear
The event ended Monday evening. The board didn’t.
Chase confirmed Monday evening that transactions were posting normally again — about 6:30 PM ET, per NBC News, with the statement reported by FOX Local at about 8:15 PM. The cause stands as Chase stated it: a software bug in on-premise hardware systems. Fox Business reported an FBI official traced it to a software release rolled out earlier Monday that affected a subset of Chase’s systems, with no indication of external cyber activity; NBC News reported regulators in contact with the banks concluded the problem was contained. The confirmed event — one bank’s bug, and the slowness felt by banks on the other side of its transactions — is closed.
Downdetector’s board held 48 entries at every hourly check from Monday afternoon through 8:30 AM Tuesday. It never cleared; it rotated. The airlines came off entirely by 1:30 AM. Gaming took the small hours — PlayStation Network climbed to #2, flagged “experiencing problems,” at 4:30 AM; War Thunder reached #1 and was gone by morning. At 8:30 AM, as business hours resumed, the banks re-consolidated: Chase #3, Bank of America #5, four banks in the top ten.
At 8:30 AM iCloud debuted at #6 on the board and Apple Support at #13, both flagged “experiencing problems.” Apple’s own system status page shows no iCloud incident — every service listed operational, as read Tuesday morning — and independent monitors find iCloud up and answering normally. That is the pattern both of Monday’s gauges measured, still running on Tuesday: the reporting layer outlives the event, and it attaches to the largest names within reach.
What remains open is only the reporting. The bank event has a stated cause and a stated end; the board has neither — it has an audience. Board presence is user reports, not confirmed outages: the standing caveat, repeated on purpose.
7:45 PM ET · The gauge, second reading
Update — 7:45 PM ET
The gauge, second reading
The gauge, second reading — the event and the echo
At 7:30 PM the board’s top five reshuffled (Chase, Bank of America, YouTube, Outlook, Wells Fargo) and new names joined — Facebook, Claude AI, Discord, Paramount+. We gauged the arrivals against their baselines. The board looked universal. The gauge says otherwise: the event is still one bank, decaying — and the rest is its echo.
All figures below were read from Downdetector service-page charts (reports vs baseline) at 7:19–7:25 PM ET this evening.
Chase181×
725 reports vs baseline 4
Down from 270× at the 6:45 reading. Day peak: 4,712 at 2:04 PM — ~942× baseline.
Wells Fargo130×
260 reports vs baseline 2
Day peak: 1,068 at 2:50 PM — 534× baseline.
Zelle is 15% of its reported problems — the rail shadow, still visible.
Microsoft Outlook37×
298 reports vs baseline 8
Day peak: 1,273 at 3:20 PM — 67× baseline.
The climbers, gauged
The board ranks by raw reports, so the biggest services float upward on ordinary noise. Gauged against their own baselines, tonight’s climbers measure very differently:
YouTube · Ranked #3 on the board7.8×
404 reports vs baseline 52
404 reports sounds large until its baseline: 52.
Facebook1.4×
23 reports vs baseline 16
The first Meta property to appear on the board all day.
On the board vsDowndetector’s own status: “no current problems.”
Claude AI2.7×
19 reports vs baseline 7
Threshold noise.
Two waves were visible at 6:45. By the second reading there is one event, decaying at its center, and a country still checking its apps. Multiples measure reporting, not confirmed failure — the same caveat as the first gauge, repeated on purpose.
6:45 PM ET · The gauge
Update — 6:45 PM ET
The gauge
How far above normal is this? Measured against Downdetector’s own baseline for each service — the typical report volume for that time of day, shown on each service page’s chart. Latest chart points, ~6:20 PM ET:
United Airlines430×
430 reports vs baseline 1
Chase270×
1,083 reports vs baseline 4
Day peak: 4,712 at ~2 PM — over 1,000× baseline.
US Bank266×
266 reports vs baseline 1
Delta Air Lines117×
117 reports vs baseline 1
Quest Diagnostics82 reports
82 reports vs baseline 0 — no multiple to compute.
A service that normally receives essentially no reports is suddenly on the board.
The discrepancy, set in type
Bank of America is the day in one frame: the reports say one thing, the bank’s own alert says the other.
The reports · ~6:20 PM137×
549 reports against a baseline of 4 — about 137× normal reporting for Bank of America.
vs
The official word · Bank of America’s own alert
operating normally today
Both can be on the page at once — because the multiple measures reporting, not a confirmed BofA failure.
PlayStation Network
On the board, flagged as having problems vsSony’s status page: all-green, all services up and running.
Amazon Web Services
On the board vsno AWS acknowledgment — no incident recognised by AWS.
Chase
“Resolve this by the end of day”vs no resolution posted yet.
Read the multiples carefully: they measure reporting — how far above its own baseline a service’s user-report volume is running — not confirmed failure. A service can sit at 137× its baseline while its operator reports it fully operational, and several have today.
The shape — two waves, not one
Wave one · The banksRose ~10 AM, peaked 1–2:30 PM, decaying since
The bank wave has been rolling over since early afternoon. Chase is down to about a quarter of its peak — 1,083 reports against 4,712 at ~2 PM.
Wave two · Late afternoonIgnited from flat baselines
United went vertical at ~4:30 PM (peak 1,105 at ~5:45 PM), Delta at ~5:15 PM, and Quest Diagnostics at ~5:30 PM — the first healthcare entry of the day. The evening board is not the morning event’s tail.
Board state — ~6:41 PM ET
Top 10 on Downdetector’s most-reported board, every one flagged “experiencing problems”:
Chase
Bank of America
United Airlines
Microsoft Outlook
Wells Fargo
US Bank
Microsoft 365
YouTube
American Airlines
OpenAI
Four airlines are on the board in total — United, American, Delta and Southwest.
Still open
Chase has not posted a resolution; its end-of-day target stands.
Monitoring continues this evening. If Chase or another institution posts a resolution, it will be added here.
5:20 PM ET · Second update
Update — 5:20 PM ET
Board update
Board state at 5:18 PM ET: still 48 services on Downdetector's most-reported list, and its shape has shifted again. Wells Fargo has moved into the top five (Chase, Bank of America, Microsoft Outlook, US Bank, Wells Fargo). United Airlines climbed from #37 to #7 in half an hour, and American Airlines has joined at #32 — two airlines are now on the board. PlayStation Network is on the board at #41, flagged “possible problems”; Sony's own status page says all services are up and running, so that listing is user reports, not an acknowledged outage. X (Twitter) has joined at #43. Amazon Web Services sits at #38, still “possible problems,” with no AWS incident acknowledged. NetSuite has dropped off the board entirely, consistent with its incident resolving at 4:12 PM.
No new official explanations have been published since the 4:55 PM update, and Chase has not yet posted a resolution; its stated target is resolution by the end of the day today.
Monitoring continues this evening. If Chase or another institution posts a resolution, it will be added here.
4:55 PM ET · First update
Update — 4:55 PM ET
Explanations on record
The explanations on record so far:
Chase — the only institution to name a cause. In its own post (3:57 PM ET), Chase said it “experienced a temporary issue related to a software bug in on-premise hardware systems” that made transactions take longer than usual. Customers still have access to their funds. Chase said the issue is not cyber related and “not a systemic issue,” and that other banks “may have been impacted by slowness in transactions only to the extent that they were a party to a transaction with us.” Chase said it is working to resolve the issue by the end of the day today.
Zelle (Early Warning Services): “Zelle is not down” — an issue at the financial institutions may be impacting broader services.
US Bank: its systems are not directly impacted; the delays are in Zelle transactions involving other institutions.
Wells Fargo: confirmed its online banking was temporarily unavailable. It has offered no cause.
Bank of America declined to comment on any outage of its own, but posted a message on Downdetector saying Zelle payments at Bank of America are operating normally — “If using another service provider and experiencing delays, please check there.” Capital One has still issued no statement as of this update.
Board state at ~4:50 PM ET: still 48 services on Downdetector's most-reported list. The top of the board (Chase, Bank of America, Microsoft Outlook, US Bank) has not changed since mid-afternoon, and the primary report curves are past the 1–2 PM peak and rolling over. Reports have since appeared for Amazon Web Services, PlayStation Network, United Airlines, Workday, T-Mobile and Xfinity — a listing is user reports, not a confirmed outage, and as of this update neither AWS nor Sony had acknowledged any incident.
NetSuite: NetSuite's separate incident (US Phoenix) was resolved at 4:12 PM ET; Oracle said it affected password sync for a subset of customers. It was never part of the bank event.
We will update this page later this evening as institutions publish explanations or resolutions.
48Services on Downdetector “Most reported” at 4:13 PM and again at 4:16 PM (parent crawls)
28+Financial brands across the 4:13 / 4:16 PM boards — banks, Zelle, Visa, PayPal, Venmo & others
9:53 AMChase reports began on Downdetector (AP / wires). Zelle reports began ~9:30 AM (USA TODAY)
>4,700Chase reports at ~2 PM (CNN). Forbes put it >5,000 around 2 PM — tallies differ by snapshot
How to read every number on this page: All Downdetector volumes are user-submitted reports at a moment in time. Outlets citing the same board give different totals because they snapshot at different minutes. No bank released its own affected-customer count. Where the research found conflicting tallies, both are shown.
What customers actually hit
Chase customers
Failed / slow logins to online and mobile banking
Balances and recent transactions not showing or updating — about half of Chase reports were balances / transactions (one tally)
Blocked wire / transfer screens; transactions taking longer than usual
Some Downdetector comments mentioned in-person systems / ATMs and cashing checks — not confirmed by Chase
“Some services aren't available. We're working on them. Please try back later.” — Chase in-app message, cited by AP
Chase stressed accounts and funds remained accessible. No lost funds were reported by any bank.
Everyone else
Wells Fargo: Online banking unavailable per its own notice — ATMs, branches and phone banking remained available
Other banks: Mainly Zelle sends / receives involving Chase — and generally other financial institutions — delayed
Zelle: 69% of Zelle reports were funds-transfer problems
Cards & rails: No confirmed impact — Visa / Mastercard / Fedwire / FedACH / FedNow showed no official incident (see layer check below)
Sources: AP, CNN, USA TODAY, Wells Fargo notice (still up at 3:18 PM ET check)
Report volume — the 2:30 PM snapshot vs. the peak tallies
Two honest views of the same board. The bars use the single 2:30 PM snapshot (App.com / Downdetector). The peak-window tallies in the bank cards below come from different minutes, so they don't match — that's the snapshot problem, not an error to hide.
The 2:30 PM spike — at a glanceCHASE IS 47× TD BANK
This is the entertaining part: scrub through Monday and watch the board change colour. Each stop is a documented moment from the timeline below — not a made-up live feed. Tap any tile to filter the bank list below to that bank.
8:00 AM — The trickle
8 AM10 AM12:48 PMPEAK 1–2 PM2:30 PM4:13 PM
Microsoft Store reports begin
Per Downdetector's own post — a trickle, not a spike. The banks aren't on the board yet. That's why “it started at 2:30” doesn't fit: the frame was already moving in the morning.
Test the theories
Pick a theory. The meter shows how much of it the official record actually supports — statements, status pages, and wires checked by ~4:40 PM.
Supported by the recordSTRONG
Chase said “technical issue, not cyber related.” Plaid flagged Chase only from 10:11 AM. Zelle and U.S. Bank both pointed back to transactions with other institutions — especially Chase. AP and CNN landed on the same shape.
Quick check — why can an ISP make every bank look down?
Tap an answer:
Timeline — it started in the morning
All times Eastern, Mon Nov 16 2026 unless noted. The 4 PM board was the tail of an event that began before 10 AM — and the board itself was already photographed, essentially complete, at 2:14 PM.
Separate Microsoft 365 outage begins
Outlook / Teams / Copilot / SharePoint. Peaked at ~130,000 Outlook reports. Microsoft later declared it fully restored Sat Nov 14.
Microsoft preliminary report; separate AWS incident
Microsoft blamed a latent defect from a July 2026 gateway fix. AWS us-east-1 had its own ~3-hour incident Nov 15, resolved — separate from today.
Microsoft Store reports begin
Per Downdetector's own post. Trickle, not a spike.
Zelle reports begin
USA TODAY live / Downdetector.
Chase problems begin on Downdetector
AP / wires citing Downdetector.
Plaid posts Chase disruption
“Chase is currently experiencing downtime impacting all channels and data recipients.” Partial outage for Chase connections only. Still open at 3:06 PM.
Renewed Microsoft reporting spike
Windows Report: Microsoft 365 >400 reports (baseline 23), Outlook >700 (baseline 32). Microsoft did not acknowledge it; official health showed operational.
Fidium Fiber problems begin
Per Downdetector. Regional.
Chase Support on X acknowledges
Working to resolve an issue with online banking; try again later.
Optimum problems begin
Local tracker — NY metro: Brooklyn / NYC / Newark / Queens / Hoboken.
Bank reports peak — early afternoon, not 4 PM
Chase 4,437 → >4,700 (Forbes >5,000); Bank of America ~1,500–1,900; U.S. Bank >900; Wells Fargo ~800; Zelle peaks ~2 PM.
Capital One flagged — and Microsoft Teams flagged
Same timestamp in Downdetector posts, two different services. No Capital One statement was found.
The full board is already there
Screenshots show essentially the same “Most reported” board: banks + Outlook / 365 / Store / Teams + Zelle + Optimum + Google / Gmail + Roblox + NetSuite + Visa. A 2:47 PM scroll showed an even longer tail (Steam, T-Mobile, Xfinity, Cloudflare, AWS).
NetSuite problems begin
Downdetector. Oracle later confirmed a separate Phoenix UI investigation at 3:11 PM.
Still underway, no resolution
Check found Wells Fargo notice still up, Plaid incident still open, no resolution from Chase. Fed status normal.
AP updated story lands the shape
Chase-centred ripple; Bank of America & Citigroup not having own-service issues; Zelle network fully operational.
Board crawl: 48 services, Chase #1
Parent crawl. Curve notes: Roblox steepest / rising; Zelle peaked & easing; Capital One still climbing; Optimum spiked & easing. Verified live at the same minute: Google Workspace — “No incidents.”
Board re-read: still 48, top 8 unchanged
Optimum rose to #9, Capital One fell to #10. New at threshold: Xbox Network #38, Spectrum #47, AT&T #48. Dropped: SoFi, Microsoft Copilot, Frontier. Roblox status verified live: “All Systems Operational.”
Microsoft verified live: no Azure issues
Service-health page showed no Microsoft 365 incident and “No Azure issues detected.”
Research cut-off — still unresolved on the record
No resolution statement from Chase, Wells Fargo or Plaid had been found. Recheck them before acting on status.
Mint ledger — the banks
Every bank, one by one
Filter by what the institution itself said — not by how high it ranked on a reports board. “No statement found” means the research found no official confirmation for Nov 16; it does not mean nothing happened to that bank's customers, especially via Zelle.
Showing all 29
Chase / JPMorgan Chase
Board #1
● Confirmed outage
“Due to a technical issue, some transactions are taking longer than usual to complete but customers still have access to their accounts and funds. We are working to resolve this as quickly as possible.” — Spokeswoman Stephanie Gostomski / Chase, to AP, CNN, USA TODAY, CBS
Also stated: “This is a technical issue, not cyber related” — a “bug,” not a cyberattack. Chase Support on X at 12:48 PM acknowledged an online-banking issue. Plaid listed a partial outage for Chase connections from 10:11 AM, still open at 3:06 PM.
Reports: >4,700 at ~2 PM (CNN) · 4,554 at 2:30 PM · >5,000 ~2 PM (Forbes) · Began 9:53 AM · Peak ~1–2 PM
“Online banking is temporarily unavailable… We're having technical issues, but we're working to resolve them as quickly as possible. Please try again later… You can still access your accounts through our ATMs, banking locations, and customer service.” — Notice on wells fargo site, still up at 3:18 PM check
The partial exception: Wells Fargo confirmed a problem of its own. No cause was given, and no link to Chase was stated. Whether its failure shares Chase's cause could not be verified.
Reports: 800 at 1:50 PM · 1,058 at 2:30 PM
Source: Wells Fargo website notice, via research notes
Zelle / Early Warning Services
Board #8
● Said network is up
“Zelle is not down. Zelle is operating as it should… An issue at the financial institutions may be impacting broader services, including Zelle payments, for some of their customers.” — Early Warning Services, to CNN, NY Post, Fox Business, NBC
NBC version: “This is a Chase issue.” AP: Zelle network was fully operational. Zelle holds no consumer funds — customers use it through their banks. The ripple ran bank → Zelle reports, not the other way.
Reports: >500 by early afternoon · 69% funds transfers · Began ~9:30 AM · Peaked ~2 PM, easing by 4:13 PM
“U.S. Bank systems are not directly impacted. However, clients attempting to send or receive Zelle payments with another financial institution may experience transaction delays.” — U.S. Bank, to CNN / AP / Gizmodo
Still #4 on the board at 4:13 and 4:16 PM — on report volume, while saying its own systems were fine. That gap is the story of this board.
No public statement as of 2:16 PM (24/7 Wall St); declined to comment (NBC Chicago). AP's updated story then stated Bank of America was not having issues with its own banking services — the impact was its customers trying to transact with Chase. Treat early “BofA outage” headlines as report volume, not a confirmed BofA failure.
Reports: ~1,100 at 1:25 PM · >1,900 at 1:50 PM · 2,097 at 2:30 PM · ~1,500 at 2 PM (Forbes) — tallies conflict by snapshot
AP's updated story named Citigroup alongside Bank of America: not having issues with its own banking services on Monday. Both Citi and Citibank appeared as separate Downdetector brands — one institution, two board entries. No separate Citi statement was found.
Downdetector flagged problems from 1:41 PM. A mirror of user-report data showed a late, fast rise — 1,238 reports in the last hour vs. ~8 usual, busiest 367 / 15 min at crawl time — consistent with it still climbing at 4:13 PM. But no bank confirmation and no cause. It fell from #9 to #10 at 4:16 PM, overtaken by Optimum.
Capital One is a Zelle / Early Warning owner-bank, so Zelle-to-Chase delays are a plausible channel — unproven for Capital One specifically.
Mirror total: 1,807 reports / 24 h at crawl · No reliable news tally
PNC
Board #15
● No statement found
On the 4:13 PM board. No official confirmation found for Nov 16 beyond the Downdetector listing.
Navy Federal
Board #16
● No statement found
On the board. A Zelle participant — the documented ripple channel — but no official confirmation found.
USAA
Board #17
● No statement found
On the board. A Zelle participant, but no official confirmation found for Nov 16.
Truist
Board #18
● No statement found
On the board. A Zelle participant, but no official confirmation found. No Truist statement was found in the research.
TD Bank
Board #21
● No statement found
On the board. The one small hard tally in the set: 96 reports in the 2:30 PM snapshot (App.com / Downdetector) — two orders of magnitude below Chase at the same minute.
JP Morgan
Board #23
● No separate statement
Listed separately from Chase on Downdetector, but it's the same banking group — Chase is JPMorgan Chase's consumer bank. No separate JP Morgan confirmation beyond Chase's.
Visa
Board #25
● No incident acknowledged
Visa Acceptance Solutions official status: Operational at 19:25 UTC (scope note: that dashboard covers the acceptance / APIs platform, not all of VisaNet). User reports did spike from ~1:01 PM on one mirror — 377 reports / 24 h, cards / prepaid top categories — with no Visa acknowledgment found. No evidence card transactions broadly failed.
Fidelity
Board #28
● No statement found
On the board. No official confirmation found for Nov 16 beyond the Downdetector listing.
Huntington
Board #29
● No statement found
On the board. No official confirmation found.
American Express
Board #30
● No statement found
On the board. No official confirmation found.
Chime
Board #31
● No statement found
On the board. No official confirmation found.
Fifth Third
Board #32
● No statement found
On the board. No official confirmation found.
PayPal
Board #33
● No statement found
On the board. No official confirmation found.
Regions
Board #34
● No statement found
On the board. No official confirmation found.
Charles Schwab
Board #35
● No statement found
On the board. No official confirmation found.
Ally
Board #36
● No statement found
On the board. No official confirmation found.
SoFi
#38 at 4:13 → off board
● No statement found
On the 4:13 PM board at #38, dropped off by 4:16 PM. No official confirmation found. Dropping off in three minutes is threshold churn, not a recovery announcement.
ADP
Board #44
● No statement found
Payroll / HR platform, on the board at #44. No official Nov 16 incident found.
Venmo
Board #45
● No statement found
On the board. No official confirmation found.
M&T Bank
Board #46
● No statement found
On the board. No official confirmation found.
KeyBank
Board #47 at 4:13
● No statement found
On the 4:13 PM board at #47 — near the threshold where the smallest report volumes still rank. No official confirmation found.
Mastercard & Fiserv — checked, not on the top board
Rails / processor
● Operational — per trackers
Mastercard: Operational at 20:20 UTC, 280–281 components up, no active / recent incidents (tracker reading its official page). Fiserv (bank processor, seen in a 2:14 PM scroll): StatusGator — up, 0 reports / 24 h. Neither supports a rails or processor cause.
The best-supported explanation is a composite of four things at once — not a single victim-producing layer. Each one is needed to fit the official timelines; a cloud / DNS / BGP mega-event fits none of them.
1 — One real failure with real wires out of it
Chase is the largest Zelle bank. Interbank P2P is the one documented shared layer in this story: a Chase slowdown mechanically produces simultaneous Zelle + partner-bank reports, and Plaid's Chase-only outage made fintech apps show Chase as down too. CNN and the updated AP both describe the other-bank reports as customers trying to transact with Chase, especially via Zelle.
2 — Reporting contagion after a brutal week
Three major US disruptions in five days: Microsoft Nov 12–14, an FAA ground stop Nov 13, banks Nov 16. News and viral screenshots from ~1 PM drove people to Downdetector, where the “most reported” ranking itself amplifies every listed service — and the threshold is low enough that a service at its usual report level (Xbox, see below) can enter at #38.
3 — Monday-afternoon peak load
Roblox, YouTube, Microsoft Store and Xbox all peak late in the school / work day. Steep late-day report curves with green official statuses fit peak use, not failure.
4 — A handful of small, separate real incidents
NetSuite's Phoenix UI (from 3:11 PM), regional Optimum (NY metro) and Fidium trouble, a minor Zoom Calendar degradation from 12:17 PM. Real — and unrelated on current evidence.
The shared-layer check — layer by layer
Only three layers sit in front of banks + Microsoft + Google + Roblox + ISPs at once: a major cloud, a major CDN / DNS provider, or backbone routing / a large ISP. Every one was checked against its official timeline for Nov 16.
Layer
Finding for Nov 16
Verdict
Microsoft Azure
All-clear all day in trackers reading the official feed; verified live ~4:18 PM: “No Azure issues detected at this time.”
No
AWS
IsInternetUp Nov 16 roundup: all clear. The us-east-1 incident in results was Nov 15 (~3 h, resolved). Elevated user reports in the afternoon, no official Nov 16 incident.
No confirmed cause
Microsoft 365
The confirmed outage was Nov 12–14 (see timeline), declared fully restored Nov 14; preliminary report Nov 15 blamed a latent July gateway defect; final report due Nov 20. On Nov 16: ~400–700 reports vs. ~130,000 at the Nov 12 peak, no admin-centre incident, no @MSFT365Status acknowledgment, official feeds operational. Elevated reporting in last week's shadow — not a documented day-long degradation, and no source links it to Chase.
Not supported
Cloudflare (CDN / DNS)
Only a narrow incident: “Changes to Cloudflare Workers are not producing Event Subscription events” (since Nov 10). DNS / CDN components listed Operational on mirrors.
Not a network-wide outage
Akamai (CDN)
One incident: Application Load Balancer reporting availability — started 10:49 PM ET Nov 15, mitigated 8:37 AM ET, resolved 9:36 AM ET. Predates the bank peak.
No
Zelle operator
Explicitly fully operational (above). Ripple ran bank → Zelle.
No — reverse direction
Visa / Mastercard
Mastercard operational, no incidents. Visa acceptance platform Operational at 19:25 UTC; user reports spiked, no acknowledgment.
No confirmed outage
Fedwire / FedACH / FedNow
No notifications, normal operations at checks ~12:30 PM, 2:35 PM and 3:18 PM ET; StatusGator: up at 18:18 UTC.
No
Fiserv (processor)
StatusGator: up, 0 reports / 24 h.
No
Plaid
Real, but Chase-specific and downstream of Chase (from 10:11 AM). Explains fintech apps showing Chase down; explains nothing else.
Effect, not cause
NetSuite (Oracle)
Genuine separate incident: “investigating an issue affecting the NetSuite Application UI in US Phoenix” from 3:11 PM ET.
Separate
ISP / backbone / BGP
No BGP, DNS or cable incident found; Nov 16 roundup: “the internet is working normally.” Regional trouble only: Optimum (NY metro, from ~1:20 PM, tens of reports / 15 min), Fidium (from 12:17 PM), Frontier (morning, CA / TX), Spectrum (minor, NY / OH, morning).
Contributor to board breadth — not the common cause
Why the ISP layer still matters to the picture: When a user's own access network fails, every service looks down to that user — and they report all of them. That mechanism manufactures board breadth without any service failing. It cannot explain the core event: bank reports began 9:30–9:53 AM, peaked in the thousands by 1–2 PM, and Chase / Plaid / Zelle confirm a bank-side mechanism independent of any ISP.
The rest of the 48 — non-banks on the board
Microsoft — elevated reports, no acknowledged outage
Outlook #3, Microsoft 365 #6, Store #12, Teams #14 at 4:13 PM. Windows Report (11:27 AM): Outlook >700 reports vs. baseline 32, M365 >400 vs. 23 — about 20× baseline, but ~200× smaller than the Nov 12 peak. Downdetector posts: Store reports since 8:00 AM, Teams since 1:41 PM. Microsoft acknowledged nothing on Nov 16; verified live ~4:18 PM, its health page showed no M365 incident. Copilot dropped off the board at 4:16 PM.
YouTube #11, Google #19, Gmail #24. Google Workspace dashboard verified live at 4:13 PM (21:13 UTC): “No incidents,” Gmail and all Workspace services Available. Independent YouTube checks: up, HTTP 200, no visitor reports in the past hour. YouTube publishes no separate public status page — that gap is noted, not filled.
Board #7, steepest curve at 4:13 PM. Verified live ~4:16 PM: “All Systems Operational.” A mirror did show a real reporting spike — 696 reports in the last hour vs. ~144 usual (2,542 / 24 h) — so users were reporting problems Roblox did not confirm. Afternoon is Roblox's peak-use window.
At 4:16 PM the carrier set was five deep: Optimum #9, Fidium Fiber, Verizon, Spectrum #47 (new), AT&T #48 (new) — Frontier dropped off. No official Nov 16 afternoon incident was found for Spectrum, AT&T, Verizon or Fidium. Entering at #47–48 means crossing the board's threshold — the lowest volumes that still rank. Xbox Network entered at #38 at the same moment with reports at its usual level (42 in the last hour vs. ~45 usual) and a green official status. Breadth widening at the bottom while the top 8 sat unchanged fits threshold churn and cross-reporting — not a new backbone failure.
Also on the board, no incident found
Amazon #20, OpenAI #27, Zoom #41, eBay #42, Zendesk #48 (at 4:13). OpenAI and Zoom were all-clear in the Nov 16 infrastructure roundup (Zoom except planned maintenance, plus that minor Calendar degradation). No official Nov 16 incidents were found for Amazon, eBay or Zendesk. NetSuite #13 is the exception — a real, separate confirmed incident (above).
No Facebook, Instagram, WhatsApp or Messenger appeared anywhere in the 48 at 4:13 or 4:16 PM. Whatever produced this board, Meta's stack was not in the set.
On the “one attack” read
The shape-match to a rolling campaign is understandable: Microsoft simmering from 8 AM, Teams flagged at 1:41 PM, banks cresting in the afternoon reads like targets taken in sequence. But on the record assembled by ~4:40 PM, it cannot be promoted to established:
Chase itself said “not cyber related”; CNBC (via multiple summaries) reported no indication of bad actors in Chase's problem and no indication of a bug hitting multiple banks at once. Reuters (via USA TODAY Network): “no indication at the time that the same bug was affecting multiple banks.”
No institution attributed anything to an attack, nobody claimed responsibility credibly, and no US agency (FDIC, CISA, Treasury) had issued any statement about today's outages in sources checked.
The sequence also dissolves on inspection: the bank event began 9:30–9:53 AM (before Teams' 1:41 PM flag) and peaked 1–2 PM; the Microsoft numbers are ~200× smaller than last week's real outage, with green official feeds throughout.
If any bank or vendor attributes this, that receipt changes the read. As of the cut-off, it doesn't exist.
Alleged / unverified — do not treat as fact
Iran claim of responsibility — viral; an OSINT analysis at 3:15 PM ET states it is not supported by any evidence available today; no agency has attributed the disruption to any actor. Ogun Security analysis
“n0n” extortion-group claim — alleged data theft from Visa / Mastercard / JPMorgan; none of the three companies has confirmed a breach, and the poster itself noted outage and data-theft claims are distinct.
Fedwire-upgrade claim — a social-media claim that a Nov 16 Fedwire release change caused this; the Fed's own communications list says the Fedwire November 2026 release was rescheduled to November 2027.
“Nov 17” rumour — wiped balances / months-long internet shutdown; explicitly debunked in the CandidDragon check. FDIC insurance covers bank failure, not app outages, and no bank reported missing funds.
Status as of ~4:40 PM ET — and what's still open
Unresolved on the record
No resolution or all-clear from Chase, Wells Fargo or Plaid in any source checked. Latest explicit check 3:18 PM: outages still underway, Wells Fargo notice still up, Plaid incident still open (listed open at 3:06 PM). USA TODAY live at 3:38 PM still described banks as experiencing difficulties.
NetSuite Phoenix UI: investigating from 3:11 PM, no resolution in sources checked.
Resolved elsewhere: Akamai ALB reporting (9:36 AM ET); Microsoft Nov 12–14 event (Nov 14, final report due Nov 20).
Could not verify — open questions
Downdetector's own live curves at 4:13–4:16 PM (the site bot-walled the research browser). Board order and curve notes rest on the parent agent's screenshots / crawls plus mirror data.
Whether Wells Fargo's failure shares Chase's cause — Wells Fargo gave no cause.
Any Bank of America / Capital One / Truist / PNC official statement — none found; BofA affirmatively declined to comment.
Exact Downdetector totals — outlets differ by snapshot (Chase 3,300–>5,000 across tallies).
Anything after ~4:40 PM ET — research cut off then. Recheck Chase / Wells Fargo / Plaid for resolution before acting.