On Thursday, three of the biggest names in payments had to explain a breach that happened to someone else.
JPMorgan Chase, Visa, and Mastercard said a cyberattack on a vendor exposed some customer information. "Based on our investigation so far," a JPMorgan spokesperson told Reuters, "we believe the attackers accessed certain marketing data held by those vendors, which may include some customer names and email addresses."
The bank said it would contact affected customers directly, and it stressed that the incident isn't related to the outage it faced earlier this week. The timing invites speculation anyway.
"One thing analysts learn early in their careers is that time is the weakest correlation," said James Shank, Director of Applied Threat Research at Expel. Customers primed by a run of outages will naturally read more into this news, he noted, even though JPMorgan has said the two aren't connected.
According to The Wall Street Journal, a Visa spokesman said the breach stemmed from a vendor operating in Asia-Pacific and that Visa's own systems played no part. Mastercard said the exposure involved a "limited amount of information from a subset of cardholders enrolled in a specific benefit program," adding that "there is no impact to the Mastercard payment network or the ability for people to use their cards."
None of the three companies has identified which vendors were breached. They also haven't disclosed how many customers were affected or whether all three firms were hit through the same systems.
According to the Journal, citing cybersecurity firm SOCRadar, a group called n0n claimed responsibility. On November 16, n0n claimed to have obtained cardmember engagement records that included "hundreds of thousands of campaign interactions for tens of thousands of distinct cardmember email addresses." When its 48-hour deadline passed, it published the data.
Cybernews researchers who examined the release counted around 16,000 customer email addresses, along with phone numbers and JPMorgan Chase account IDs. The data appear to relate to the bank's Reserve program, an invitation-only Visa credit card that usually requires customers to have at least $10 million in assets under management. The researchers also found more than 100 employee email addresses, and they counted records on more than 6,000 Visa customers.
The released JPMorgan dataset was only 502 KB, and it was unclear whether it represented a sample or everything the attackers had stolen. JPMorgan has not named the vendors or confirmed n0n's responsibility or the 16,000 figure. The alleged Mastercard records are even less certain. Researchers say the 2,000 entries may be test data, and their authenticity is still unverified.
Names and email addresses rarely make headlines. This list is different because of who is on it. A verified roster of ultra-high-net-worth clients, tied to one premium card program and sometimes paired with account IDs, is a ready-made target list for whaling and impersonation.
"The marketing information may be rich enough to allow targeted phishing attacks," Shank said, "and may include enough details to make it look very real." He sees a window for JPMorgan to warn customers about the heightened risk while its threat intelligence team watches whether the data start showing up in scams.
The incident also points to a common blind spot in vendor tiering. Marketing, loyalty, and benefits platforms often sit in lower-risk tiers because they don't touch card numbers or core banking systems. Even when a vendor is tiered correctly, leverage is limited. "Third-party risks are still an unsolved problem within cybersecurity," Shank said, and few organizations have any real way to control their suppliers' security posture. Whether data are sensitive depends on context as well as data type. The same email field is low-value on a newsletter list and high-value on a list of people with eight-figure balances.
Several other incidents this month have followed the same pattern, where the organization customers trust wasn't the one that got breached.
Irish credit unions – Attackers exploited a flaw in Wellington IT's online banking platform between October 2 and 4 and took saved payee records, including names, account numbers, and IBANs. More than 40 credit unions didn't hear about it until November 16, a delay Minister of State Robert Troy called "not acceptable." Wellington IT holds a 51% share of the Irish market, so for half the sector, one supplier became a single point of failure.
Optimizely – An attacker abused a flaw in Optimizely’s Web Experimentation signup flow to plant malicious JavaScript in customer projects. Because the code was still served from Optimizely's CDN, sites for 178 organizations, including HSBC, Zoom, and the BBC, delivered card skimmers and crypto-stealers to visitors without being breached themselves. The few sites with restrictive Content Security Policy rules blocked it.
[RELATED: OPTIMJACK Supply-Chain Attack Exposes Visitors Across 178 Major Brands]
Notion and Fireworks AI – Notion's incident notice shows how a vendor's account can change. Fireworks first reported that only synthetic test data were exposed, then confirmed six days later that real Notion AI conversations were included, affecting fewer than 1% of workspaces. The data had reached Fireworks through an analytics feature, not a headline integration.
Inventory the vendors that sit off the critical path. Third-party programs tend to focus on payment processors, cloud hosts, and core platforms. This week's incidents came through marketing and loyalty vendors, a benefits program, an online banking supplier, an A/B testing script, and an AI subprocessor. If a vendor holds customer data or runs code in customers' browsers, it belongs in scope.
Tier by blast radius, not just data type. Ask who is in the dataset and how many of your peers share the same supplier. A vendor serving half a market or holding your most valuable clients needs top-tier scrutiny, even if its data look mundane.
Write notification timelines into contracts, and expect early findings to change. About six weeks passed between the Wellington IT intrusion and the credit unions hearing about it. Notion's first answer was "synthetic data," and that changed within a week. Contracts should require defined notification windows and scope, and incident communications should be written so they can be updated without contradicting earlier statements.
Treat third-party JavaScript, and your vendors' account controls, as part of your attack surface. The Optimizely attackers never touched customers' servers; they abused the vendor's signup flow. A restrictive CSP on script-src and connect-src, plus monitoring for changes to trusted scripts, separated the protected sites from the infected ones.
Get ahead of the follow-on fraud. Contact and payee data tend to resurface as impersonation and payment-redirection scams. Customer notices should spell out what the organization will never ask for.
The companies that were breached this week mostly weren't the ones customers deal with. Customers will still hear about it from their bank, card network, or credit union, so the job of explaining it falls on those brands.