Amid the hustle and bustle of the RSA Security Conference in San Francisco last week, researchers at RSA released a startling report that received very little press coverage relative to its overall importance. The report detailed a malware campaign that piggybacked on a popular piece of software used by system administrators at some of the nation’s largest companies. Incredibly, the report did not name the affected software, and the vendor in question has apparently chosen to bury its breach disclosure. This post is an attempt to remedy that.

The RSA report detailed the threat from a malware operation the company dubbed “Kingslayer.” According to RSA, the attackers compromised the Web site of a company that sells software to help Windows system administrators better parse and understand Windows event logs. RSA said the site hosting the event log management software was only compromised for two weeks — from April 9, 2015 to April 25, 2015 — but that the intrusion was likely far more severe than the short duration of the intrusion suggests.

That’s because in addition to compromising the download page for this software package, the attackers also hacked the company’s software update server, meaning any company that already had the software installed prior to the site compromise would likely have automatically downloaded the compromised version when the software regularly checked for available updates (as it was designed to do).

RSA said that in April 2016 it “sinkholed” or took control over the Web site that the malware used as a control server — oraclesoft[dot]net — and from there they were able to see indicators of which organizations might still be running the backdoored software. According to RSA, the victims included five major defense contractors; four major telecommunications providers; 10+ western military organizations; more than two dozen Fortune 500 companies; 24 banks and financial institutions; and at least 45 higher educational institutions.

RSA declined to name the software vendor whose site was compromised, but said the company issued a security notification on its Web site on June 30, 2016 and updated the notice on July 17, 2016 at RSA’s request following findings from further investigation into a defense contractor’s network. RSA also noted that the victim software firm had a domain name ending in “.net,” and that the product in question was installed as a Windows installer package file (.msi).

Using that information, it wasn’t super difficult to find the product in question. An Internet search for the terms “event log security notification april 2015” turns up a breach notification from June 30, 2016 about a software package called EVlog, produced by an Altair Technologies Ltd. in Mississauga, Ontario. The timeline mentioned in the breach notification exactly matches the timeline laid out in the RSA report.

As far as breach disclosures go, this one is about the lamest I’ve ever seen given the sheer number of companies that Altair Technologies lists on its site as subscribers to eventid.net, an online service tied to EVlog. I could not locate a single link to this advisory anywhere on the company’s site, nor could I find evidence that Altair Technologies had made any effort via social media or elsewhere to call attention to the security advisory; it is simply buried in the site. A screenshot of the original, much shorter, version of that notice is here.

Perhaps the company emailed its subscribers about the breach, but that seems doubtful. The owner of Altair Technologies, a programmer named Adrian Grigorof, did not respond to multiple requests for comment.

“This attack is unique in that it appears to have specifically targeted Windows system administrators of large and, perhaps, sensitive organizations,” RSA said in its report. “These organizations appeared on a list of customers still displayed on the formerly subverted software vendor’s Web site. This is likely not coincidence, but unfortunately, nearly two years after the Kingslayer campaign was initiated, we still do not know how many of the customers listed on the website may have been breached, or possibly are still compromised by the Kingslayer perpetrators.”

It’s perhaps worth noting that this isn’t the only software package sold by Altair Technologies. An analysis of Eventid.net shows that the site is hosted on a server along with three other domains, eventreader.com, firegen.com and grigorof.com (the latter being a vanity domain of the software developer). The other two domains — eventreader.com and firegen.com — correspond to different software products sold by Altair.

The fact that those software titles appear to have been sold and downloadable from the same server as eventid.net (going back as far as 2010) suggests that those products may have been similarly compromised. However, I could find no breach notification mentioning those products. Here is a list of companies that Altair says are customers of Firegen; they include 3M, DirecTV, Dole Food Company, EDS, FedEx, Ingram Micro, Northrup Grumman, Symantec and the U.S. Marshals Service.

RSA calls these types of intrusions “supply chain attacks,” in that they provide one compromise vector to multiple targets. It’s not difficult to see from the customer lists of the software titles mentioned above why an attacker might salivate over the idea of hacking an entire suite of software designed for corporate system administrators.

“Supply chain exploitation attacks, by their very nature, are stealthy and have the potential to provide the attacker access to their targets for a much longer period than malware delivered by other common means, by evading traditional network analysis and detection tools,” wrote RSA’s Kent Backman and Kevin Stear. “Software supply chain attacks offer considerable ‘bang for the buck’ against otherwise hardened targets. In the case of Kingslayer, this especially rings true because the specific system-administrator-related systems most likely to be infected offer the ideal beachhead and operational staging environment for system exploitation of a large enterprise.”

A copy of the RSA report is available here (PDF).

A handful of readers have inquired as to the whereabouts of Microsoft‘s usual monthly patches for Windows and related software. Microsoft opted to delay releasing any updates until next month, even though there is a zero-day vulnerability in Windows going around. However, Adobe did push out updates this week as per usual to fix critical issues in its Flash Player software.

In a brief statement this week, Microsoft said it “discovered a last minute issue that could impact some customers” that was not resolved in time for Patch Tuesday, which normally falls on the second Tuesday of each month. In an update to that advisory posted on Wednesday, Microsoft said it would deliver February’s batch of patches as part of the next regularly-scheduled Patch Tuesday, which falls on March 14, 2017.

On Feb. 2, the CERT Coordination Center at Carnegie Mellon University warned that an unpatched bug in a core file-sharing component of Windows (SMB) could let attackers crash Windows 8.1, and Windows 10 systems, as well as server equivalents of those platforms. CERT warned that exploit code for the flaw was already available online.

The updates from Adobe fix at least 13 vulnerabilities in versions of Flash Player for Windows, Mac, ChromeOS and Linux systems. Adobe said it is not aware of any exploits in the wild for any of the 13 flaws fixed in this update.

The latest update brings Flash to v. 24.0.0.221. The update is rated “critical” for all OSes except Linux; critical flaws can be exploited to compromise a vulnerable system through no action on the part of the user, aside from perhaps browsing to a malicious or hacked Web site.

Flash has long been a risky program to leave plugged into the browser. If you have Flash installed, you should update, hobble or remove Flash as soon as possible. To see which version of Flash your browser may have installed, check out this page.

The smartest option is probably to ditch the program once and for all and significantly increase the security of your system in the process. An extremely powerful and buggy program that binds itself to the browser, Flash is a favorite target of attackers and malware. For some ideas about how to hobble or do without Flash (as well as slightly less radical solutions) check out A Month Without Adobe Flash Player.

If you choose to keep and update Flash, please do it today. The most recent versions of Flash should be available from the Flash home page. Windows users who browse the Web with anything other than Internet Explorer may need to apply this patch twice, once with IE and again using the alternative browser (Firefox, Opera, e.g.).

Chrome and IE should auto-install the latest Flash version on browser restart (users may need to manually check for updates in and/or restart the browser to get the latest Flash version). Chrome users may need to restart the browser to install or automatically download the latest version. When in doubt, click the vertical three dot icon to the right of the URL bar, select “Help,” then “About Chrome”: If there is an update available, Chrome should install it then.

It’s been a remarkable week for cyber justice. On Thursday, a Ukrainian man who hatched a plan in 2013 to send heroin to my home and then call the cops when the drugs arrived was sentenced to 41 months in prison for unrelated cybercrime charges. Separately, a 19-year-old American who admitted to being part of a hacker group that sent a heavily-armed police force to my home in 2013 was sentenced to three years probation.

Sergey Vovnenko, a.k.a. “Fly,” “Flycracker” and “MUXACC1,” pleaded guilty last year to aggravated identity theft and conspiracy to commit wire fraud. Prosecutors said Vovnenko operated a network of more than 13,000 hacked computers, using them to harvest credit card numbers and other sensitive information.

When I first became acquainted with Vovnenko in 2013, I knew him only by his many hacker names, including “Fly” and “Flycracker,” among others. At the time, Fly was the administrator of the fraud forum “thecc[dot]bz,” an exclusive and closely guarded Russian language board dedicated to financial fraud and identity theft.

After I secretly gained access to his forum, I learned he’d hatched a plot to have heroin sent to my home and to have one of his forum lackeys call the police when the drugs arrived.

I explained this whole ordeal in great detail in 2015, when Vovnenko initially was extradited from Italy to face charges here in the United States. In short, the antics didn’t end when I foiled his plot to get me arrested for drug possession, and those antics likely contributed to his arrest and to this guilty plea.

Vovnenko contested his extradition from Italy, and in so doing spent roughly 15 months in arguably Italy’s worst prison. During that time, he seemed to have turned his life around, sending me postcards at Christmas time and even an apparently heartfelt apology letter.

On Thursday, a judge in New Jersey sentenced Vovnenko to 41 months in prison, three years of supervised released and ordered him to pay restitution of $83,368.

Separately, a judge in Washington, D.C. handed down a sentence of three year’s probation to Eric Taylor, a hacker probably better known by his handle “Cosmo the God.”

Taylor was among several men involved in making a false report to my local police department at the time about a supposed hostage situation at our Virginia home. In response, a heavily-armed police force surrounded my home and put me in handcuffs at gunpoint before the police realized it was all a dangerous hoax known as “swatting.”

CosmoTheGod rocketed to Internet infamy in 2013 when he and a number of other hackers set up the Web site exposed[dot]su, which routinely “doxed” dozens of public officials and celebrities by publishing the address, Social Security numbers and other personal information on the former First Lady Michelle Obama, the then-director of the FBI and the U.S. attorney general, among others. The group also swatted many of the people they doxed.

Taylor and his co-conspirators were able to dox so many celebrities and public officials because they hacked a Russian identity theft service called ssndob[dot]ru. That service in turn relied upon compromised user accounts at data broker giant LexisNexis to pull personal and financial data on millions of Americans.

At least two other young men connected to the exposed[dot]su conspiracy have already been sentenced to prison.

Among them was Mir Islam, a 22-year-old Brooklyn man who was sentenced last year to two years in prison for doxing and swatting, and for cyberstalking a young woman whom he also admitted to swatting. Because he served almost a year of detention prior to his sentencing, Islam was only expected to spend roughly a year in prison, although it appears he was released before even serving the entire year.

Hours after his sentencing, Taylor reached out to KrebsOnSecurity via Facetime to apologize for his actions. Taylor, a California native, said he is trying to turn his life around, and that he has even started his own cybersecurity consultancy.

“I live in New York City now, have a baby on the way and am really trying to get my shit together finally,” Taylor said.

If Taylor’s physical appearance is any indication, he is indeed turning over a new leaf. At the time he was involved in publishing exposed[dot]su, the six-foot, seven-inch CosmoTheGod was easily a hundred pounds heavier than he is now.

Unfortunately, not everyone in Taylor’s former crew is making changes for the better. According to Taylor, his former co-conspirator Islam was recently re-arrested after allegedly cyberstalking Taylor’s girlfriend. That stalking claim could not be independently confirmed, however court documents show that Islam was indeed re-arrested and incarcerated last month in New York.

Late last month, multiple news outlets reported that unspecified law enforcement officials had seized the servers for Leakedsource.com, perhaps the largest online collection of usernames and passwords leaked or stolen in some of the worst data breaches — including billions of credentials for accounts at top sites like LinkedIn, Myspace, and Yahoo.

In a development that could turn out to be deeply ironic, it seems that the real-life identity of LeakedSource’s principal owner may have been exposed by many of the same stolen databases he’s been peddling.

LeakedSource in October 2015 began selling access to passwords stolen in high-profile breaches. Enter any email address on the site’s search page and it would tell you if it had a password corresponding to that address. However, users had to select a payment plan before viewing any passwords.

LeakedSource was a curiosity to many, and for some journalists a potential source of news about new breaches. But unlike services such as BreachAlarm and HaveIBeenPwned.com — which force users to verify that they can access a given account or inbox before the site displays whether it has found a password associated with the account in question — LeakedSource did nothing to validate users. This fact, critics charged, showed that the proprietors of LeakedSource were purely interested in making money and helping others pillage accounts.

I also was curious about LeakedSource, but for a different reason. I wanted to chase down something I’d heard from multiple sources: That one of the administrators of LeakedSource also was the admin of abusewith[dot]us, a site unabashedly dedicated to helping people hack email and online gaming accounts.

Abusewith[dot]us began in September 2013 as a forum for learning and teaching how to hack accounts at Runescape, a massively multiplayer online role-playing game (MMORPG) set in a medieval fantasy realm where players battle for kingdoms and riches.

The currency with which Runescape players buy and sell weapons, potions and other in-game items are virtual gold coins, and many of Abusewith[dot]us’s early members traded in a handful of commodities: Phishing kits and exploits that could be used to steal Runescape usernames and passwords from fellow players; virtual gold plundered from hacked accounts; and databases from hacked forums and Web sites related to Runescape and other online games.

The administrator of Abusewith[dot]us is a hacker who uses the nickname “Xerx3s.” The avatar attached to Xerx3s’s account suggests the name is taken from Xerxes the Great, a Persian king who lived during the fifth century BC.

Xerx3s the hacker appears to be especially good at breaking into discussion forums and accounts dedicated to Runescape and online gaming. Xerx3s also is a major seller of Runescape gold — often sold to other players at steep discounts and presumably harvested from hacked accounts.

I didn’t start looking into who might be responsible for LeakedSource until July 2016, when I sought an interview by reaching out to the email listed on the site (leakedsourceonline@gmail.com). Soon after, I received a Jabber chat invite from the address “leakedsource@chatme.im.”

The entirety of that brief interview is archived here. I wanted to know whether the proprietors of the service believed they were doing anything wrong (we’ll explore more about the legal aspects of LeakedSource’s offerings later in this piece).  Also, I wanted to learn whether the rumors of LeakedSource arising out of Abusewith[us] were true.

“After many of the big breaches of 2015, we noticed a common public trend…’Where can I search it to see if I was affected?’,” wrote the anonymous person hiding behind the leakedsource@chatme.im account. “And thus, the idea was born to fill that need, not rising out of anything. We are however going to terminate the interview as it does seem to be more of a witch hunt instead of journalism. Thank you for your time.”

Nearly two weeks after that chat with the LeakedSource administrator, I got a note from a source who keeps fairly close tabs on the major players in the English-speaking cybercrime underground. My source told me he’d recently chatted with Xerx3s using the Jabber address Xerx3s has long used prior to the creation of LeakedSource — xerx3s@chatme.im.

Xerx3s told my source in great detail about my conversation with the Leakedsource administrator, suggesting that either Xerx3s was the same person I spoke with in my brief interview with LeakedSource, or that the LeakedSource admin had shared a transcript of our chat with Xerx3s.

Although his username on Abusewith[dot]us was Xerx3s, many of Xerx3s’s closest associates on the forum referred to him as “Wade” in their forum postings. This is in reference to a pseudonym Xerx3s frequently used, “Jeremy Wade.”

One email address this Jeremy Wade identity used pseudonymously was imjeremywade@gmail.com. According to a “reverse WHOIS” record search ordered through Domaintools.com, that email address is tied to two domain names registered in 2015: abusing[dot]rs, and cyberpay[dot]info. The original registration records for each site included the name “Secure Gaming LLC.”

The “Jeremy Wade” pseudonym shows up in a number of hacked forum databases that were posted to both Abusewith[dot]us and LeakedSource, including several other sites related to hacking and password abuse.

For example, the user database stolen and leaked from the DDoS-for-hire service “panic-stresser[dot]xyz” shows that a PayPal account tied to the email address eadeh_andrew@yahoo.com paid $5 to cover a subscription for a user named “jeremywade;” The leaked Panicstresser database shows the Jeremywade account was tied to the email address xdavros@gmail.com, and that the account was created in July 2012.

The leaked Panicstresser database also showed that the first login for that Jeremywade account came from the Internet address 68.41.238.208, which is a dynamic Internet address assigned to residential customers of Comcast Communications in Michigan.

According to a large number of forum postings, it appears that whoever used the xdavros@gmail.com address also created several variations on that address, including alexdavros@gmail.com, davrosalex3@yahoo.com, davrosalex4@yahoo.com, as well as themarketsales@gmail.com.

The Gmail account xdavros@gmail.com was used to register at least four domain names almost six years ago in 2011. Two of those domains — daily-streaming.com and tiny-chats.com — were originally registered to a “Nick Davros” at 3757 Dunes Parkway, Muskegon, Mich. The other two were registered to a Nick or Alex Davros at 868 W. Hile Rd., Muskegon, Mich. All four domain registration records included the phone number +12313430295.

I took that 68.41.238.208 Internet address that the leaked Panicstresser database said was tied to the account xdavros@gmail.com and ran an Internet search on it. The address turned up in yet another compromised hacker forum database — this time in the leaked user database for sinister[dot]ly, ironically another site where users frequently post databases plundered from other sites and forums.

The leaked sinister[dot]ly forum database shows that a user by the name of “Jwade” who registered under the email address trpkisaiah@gmailcom first logged into the forum from the same Comcast Internet address tied to the xdavros@gmail.com account at Panicstresser.

I also checked that Michigan Comcast address with Farsight Security, a security firm which runs a paid service that tracks the historic linkages between Internet addresses and domain names. Farsight reported that between 2012 and 2014, the Internet address 68.41.238.208 was tied to no-ip.biz, popular “dynamic IP” service.

No-ip.biz and other dynamic IP address services are usually free services that allow users to have Web sites hosted on servers that frequently change their Internet addresses. This type of service is useful for people who want to host a Web site on a home-based Internet address that may change from time to time, because services like No-ip.biz can be used to easily map the domain name to the user’s new Internet address whenever it happens to change.

Unfortunately, these dynamic IP providers are extremely popular in the attacker community, because they allow bad guys to keep their malware and scam sites up even when researchers mange to track the attacking IP address and convince the ISP responsible for that address to disconnect the malefactor. In such cases, dynamic IP services allow the owner of the attacking domain to simply re-route the attack site to another Internet address that he controls.

Farsight reports that the address 68.41.238.208 maps back to three different dynamic IP domains, including “jwade69.no-ip.biz,” “wadewon.no-ip.biz,” and “jrat6969.zapto.org.” That first dynamic address — jwade69.no-ip.biz — was included among several hundred others in a list published by the Federal Bureau of Investigation as tied to the distribution of Blackshades, a popular malware strain that was used as a password-stealing trojan by hundreds of paying customers prior to May 2014.

XERX3S HACKED?

In January 2017, when news of the alleged raid on LeakedSource began circulating in the media, I began going through my notes and emails searching for key accounts known to be tied to Xerx3s and the administrator of Abusewith[dot]us.

Somehow, in the previous three months I’d managed to overlook an anonymous message I received in mid-September from a reader who claimed to have hacked the email account themarketsales@gmail.com, one of several addresses my research suggested was tied to Xerx3s.

The anonymous source didn’t say exactly how he hacked this account, but judging from the passwords tied to Xerx3s’s other known accounts that were included in the various forum database leaks listed above it may well have been because Xerx3s in some cases re-used the same password across multiple accounts. 

My anonymous source shared almost a dozen screenshots of his access to themarketsales@gmail.com, which indicate the name attached to the account was “Alex Davros.” The screenshots also show this user received thousands of dollars in Paypal payments from Leakedsource.com over a fairly short period in 2015.

The screenshots also showed that themarketsales@gmail.com was tied to a PayPal account assigned to a Secured Gaming LLC. Recall that this is the same company name included in the Web site registration records back in 2011 for daily-streaming.com and tiny-chats.com.

In addition, the screenshot above and others shared by my source indicate that the same Paypal account tied to themarketsales@gmail.com was habitually used to pay a monthly bill from Hyperfilter.com, a company that provides DDoS protection and hosting and which has long been the provider used by Abusewith[dot]us.

Finally, the anonymous hacker shared screenshots suggesting he had also hacked into the email account desiparker18@gmail.com, an account apparently connected to a young lady in Michigan named Desi Parker. The screenshots for Ms. Parker suggest her hacked Gmail account was tied to an Apple iTunes account billed to a MasterCard ending in 7055 and issued to an Alexander Davros at 868 W. Hile, Muskegon, Mich.

The screenshots show the desiparker18@gmail.com address is associated with an Instagram account for a woman by the same name from Muskegon, Mich. (note that the address given in the WHOIS records for Alex Davros’s daily-streaming.com and tiny-chats.com also was Muskegon, Mich).

Desi Parker’s Instagram lists her “spouse” as Alex Davros, and says her phone number is 231-343-0295. Recall that this is the same phone number included in the Alex Davros domain registration records for daily-streaming.com and tiny-chats.com. That phone number is currently not in service.

Desi Parker’s Facebook account indeed says she is currently in a relationship with Alexander Marcus Davros, and the page links to this Facebook account for Alex Davros.

Alex’s Facebook profile is fairly sparse (at least the public version of it), but there is a singular notation in his entire profile that stands out: Beneath the “Other Names” heading under the “Details about Alexander” tab, Alex lists “TheKing.” Parker’s Instagram account includes a photo of an illustration she made including her beau’s first name with a crown on top.

Interestingly, two email addresses connected to domains associated with the Jeremy Wade alias — matt96sk@yahoo.com and skythekiddy@yahoo.com — are tied to Facebook accounts for Michigan residents who both list Alex Davros among their Facebook friends.

Below is a rough mind map I created which attempts to show the connections between the various aliases, email addresses, phone numbers and Internet addresses mentioned above. At a minimum, they strongly indicate that Xerx3s is indeed an administrator of LeakedSource.

I managed to reach Davros through Twitter, and asked him to follow me so that we could exchange direct messages. Within maybe 60 seconds of my sending that tweet, Davros followed me on Twitter and politely requested via direct message that I remove my public Twitter messages asking him to follow me.

After I did as requested, Davros’s only response initially was, “Wow, impressive but I can honestly tell you I am not behind the service.” However, when pressed to be more specific, he admitting to being Xerx3s but claimed he had no involvement in LeakedSource.

“I am xer yes but LS no,” Davros said. He stopped answering my questions after that, saying he was busy “doing a couple things IRL.” IRL is Internet slang for “in real life.” Presumably these other things he was doing while I was firing off more questions had nothing to do with activities like deleting profiles or contacting an attorney.

Even if Davros is telling the truth, the preponderance of clues here and the myriad connections between them suggest that he at least has close ties to some of those who are involved in running LeakedSource.

THE LEGALITY OF LEAKEDSOURCE

On the surface, the rationale that LeakedSource’s proprietors have used to justify their service may seem somewhat reasonable: The service merely catalogs information that is already stolen from companies and that has been leaked in some form online.

But legal experts I spoke with saw things differently, saying LeakedSource’s owners could face criminal charges if prosecutors could show LeakedSource intended for the passwords that are for sale on the site to be used in the furtherance of a crime.

Orin Kerr, director of the Cybersecurity Law Initiative at The George Washington University, said trafficking in passwords is clearly a crime under the Computer Fraud and Abuse Act (CFAA).

Specifically, Section A6 of the CFAA, which makes it a crime to “knowingly and with intent to defraud traffic in any password or similar information through which a computer may be accessed without authorization, if…such trafficking affects interstate or foreign commerce.

“CFAA quite clearly punishes password trafficking,” Kerr said. “The statute says the [accused] must be trafficking in passwords knowingly and with intent to defraud, or trying to further unauthorized access.”

Judith Germano, a senior fellow at the Center on Law and Security at New York University’s School of Law, said LeakedSource might have a veneer of legitimacy if it made an effort to check whether users already have access to the accounts for which they’re seeking passwords.

“If they’re not properly verifying that when the user goes to the site to get passwords then I think that’s where their mask of credibility falls,” Germano said.

LeakedSource may be culpable also because at one point the site offered to crack hashed or encrypted passwords for a fee. In addition, it seems clear that the people who ran the service also advocated the use of stolen passwords for financial gain.

Sources at nearly a half-dozen banks and credit unions independently reached out over the past 48 hours to inquire if I’d heard anything about a data breach at Arby’s fast-food restaurants. Asked about the rumors, Arby’s told KrebsOnSecurity that it recently remediated a breach involving malicious software installed on payment card systems at hundreds of its restaurant locations nationwide.

A spokesperson for Atlanta, Ga.-based Arby’s said the company was first notified by industry partners in mid-January about a breach at some stores, but that it had not gone public about the incident at the request of the FBI.

“Arby’s Restaurant Group, Inc. (ARG) was recently provided with information that prompted it to launch an investigation of its payment card systems,” the company said in a written statement provided to KrebsOnSecurity.

“Upon learning of the incident, ARG immediately notified law enforcement and enlisted the expertise of leading security experts, including Mandiant,” their statement continued. “While the investigation is ongoing, ARG quickly took measures to contain this incident and eradicate the malware from systems at restaurants that were impacted.”

Arby’s said the breach involved malware placed on payment systems inside Arby’s corporate stores, and that Arby’s franchised restaurant locations were not impacted.

Arby’s has more than 3,330 stores in the United States, and roughly one-third of those are corporate-owned. The remaining stores are franchises. However, this distinction is likely to be lost on Arby’s customers until the company releases more information about individual restaurant locations affected by the breach.

“Although there are over 1,000 corporate Arby’s restaurants, not all of the corporate restaurants were affected,” said Christopher Fuller, Arby’s senior vice president of communications. “But this is the most important point: That we have fully contained and eradicated the malware that was on our point-of-sale systems.”

The first clues about a possible breach at the sandwich chain came in a non-public alert issued by PSCU, a service organization that serves more than 800 credit unions.

The alert sent to PSCU member banks advised that PSCU had just received very long lists of compromised card numbers from both Visa and MasterCard. The alerts stated that a breach at an unnamed retailer compromised more than 355,000 credit and debit cards issued by PCSU member banks.

“PSCU believes the alerts are associated with a large fast food restaurant chain, yet to be announced to the public,” reads the alert, which was sent only to PSCU member banks.

Arby’s declined to say how long the malware was thought to have stolen credit and debit card data from infected corporate payment systems. But the PSCU notice said the breach is estimated to have occurred between Oct. 25, 2016 and January 19, 2017.

Such a large alert from the card associations is generally a sign of a sizable nationwide breach, as this is likely just the first of many alerts Visa and MasterCard will send to card-issuing banks regarding accounts that were compromised in the intrusion. If history is any lesson, some financial institutions will respond by re-issuing thousands of customer cards, while other (likely larger) institutions will focus on managing fraud losses on the compromised cards.

The breach at Arby’s comes as many credit unions and smaller banks are still feeling the financial pain from fraud related to a similar breach at the fast food chain Wendy’s. KrebsOnSecurity broke the news of that breach in January 2016, but the company didn’t announce it had fully removed the malware from its systems until May 2016. But two months after that the company was forced to admit that many Wendy’s locations were still compromised.

B. Dan Berger, president and CEO of the National Association of Federal Credit Unions, said the number of cards that PSCU told member banks were likely exposed in this breach is roughly in line with the numbers released not long after news of the Wendy’s breach broke.

“Hundreds of thousands of cards is a big number, and with the Wendy’s breach, the alerts we were getting from Visa and MasterCard were in the six-digit ranges for sure,” Berger said. “That’s probably one of the biggest numbers I’ve heard.”

Berger said the Wendy’s breach was especially painful because the company was re-compromised after it scrubbed its payment systems of malicious software. Many banks and credit unions ended up re-issuing customer cards several times throughout last year after loyal Wendy’s customers re-compromised their brand new cards again and again because they routinely ate at multiple Wendy’s locations throughout the month.

“We had institutions that stopped approving debit and credit transactions through Wendy’s when they were still dealing with that breach,” Berger said. “Our member credit unions were eating the costs of fraud on compromised cards, and on top of that having to re-issue the same cards over and over.”

Point-of-sale malware has driven most of the major retail industry credit card breaches over the past two years, including intrusions at Target and Home Depot, as well as breaches at a slew of point-of-sale vendors. The malware sometimes is installed via hacked remote administration tools like LogMeIn; in other cases the malware is relayed via “spear-phishing” attacks that target company employees. Once the attackers have their malware loaded onto the point-of-sale devices, they can remotely capture data from each card swiped at that cash register.

Thieves can then sell that data to crooks who specialize in encoding the stolen data onto any card with a magnetic stripe, and using the cards to purchase high-priced electronics and gift cards from big-box stores like Target and Best Buy.

Readers should remember that they’re not liable for fraudulent charges on their credit or debit cards, but they still have to report the unauthorized transactions. There is no substitute for keeping a close eye on your card statements. Also, consider using credit cards instead of debit cards; having your checking account emptied of cash while your bank sorts out the situation can be a hassle and lead to secondary problems (bounced checks, for instance).