Posts

NY Payroll Company Vanishes With $35 Million

MyPayrollHR, a now defunct cloud-based payroll processing firm based in upstate New York, abruptly ceased operations this past week after stiffing employees at thousands of companies. The ongoing debacle, which allegedly involves malfeasance on the part of the payroll company’s CEO, resulted in countless people having money drained from their bank accounts and has left nearly $35 million worth of payroll and tax payments in legal limbo.

Unlike many stories here about cloud service providers being extorted by hackers for ransomware payouts, this snafu appears to have been something of an inside job. Nevertheless, it is a story worth telling, in part because much of the media coverage of this incident so far has been somewhat disjointed, but also because it should serve as a warning to other payroll providers about how quickly and massively things can go wrong when a trusted partner unexpectedly turns rogue.

Clifton Park, NY-based MyPayrollHR — a subsidiary of ValueWise Corp. — disclosed last week in a rather unceremonious message to some 4,000 clients that it would be shutting its virtual doors and that companies which relied upon it to process payroll payments should kindly look elsewhere for such services going forward.

This communique came after employees at companies that depend on MyPayrollHR to receive direct deposits of their bi-weekly payroll payments discovered their bank accounts were instead debited for the amounts they would normally expect to accrue in a given pay period.

To make matters worse, many of those employees found their accounts had been dinged for two payroll periods — a month’s worth of wages — leaving their bank accounts dangerously in the red.

The remainder of this post is a deep-dive into what we know so far about what transpired, and how such an occurrence might be prevented in the future for other payroll processing firms.

A $26 MILLION TEXT FILE

To understand what’s at stake here requires a basic primer on how most of us get paid, which is a surprisingly convoluted process. In a typical scenario, our employer works with at least one third party company to make sure that on every other Friday what we’re owed gets deposited into our bank account.

The company that handled that process for MyPayrollHR is a California firm called Cachet Financial Services. Every other week for more than 12 years, MyPayrollHR has submitted a file to Cachet that told it which employee accounts at which banks should be credited and by how much.

According to interviews with Cachet, the way the process worked ran something like this: MyPayrollHR would send a digital file documenting deposits made by each of these client companies which laid out the amounts owed to each clients’ employees. In turn, those funds from MyPayrollHR client firms then would be deposited into a settlement or holding account maintained by Cachet.

From there, Cachet would take those sums and disburse them into the bank accounts of people whose employers used MyPayrollHR to manage their bi-weekly payroll payments.

But according to Cachet, something odd happened with the instructions file MyPayrollHR submitted on the afternoon of Wednesday, Sept. 4 that had never before transpired: MyPayrollHR requested that all of its clients’ payroll dollars be sent not to Cachet’s holding account but instead to an account at Pioneer Savings Bank that was operated and controlled by MyPayrollHR.

The total amount of this mass payroll deposit was approximately $26 million. Wendy Slavkin, general counsel for Cachet, told KrebsOnSecurity that her client then inquired with Pioneer Savings about the wayward deposit and was told MyPayrollHR’s bank account had been frozen.

Nevertheless, the payroll file submitted by MyPayrollHR instructed financial institutions for its various clients to pull $26 million from Cachet’s holding account — even though the usual deposits from MyPayrollHR’s client banks had not been made.

REVERSING THE REVERSAL

In response, Cachet submitted a request to reverse that transaction. But according to Slavkin, that initial reversal request was improperly formatted, and so Cachet soon after submitted a correctly coded reversal request.

Financial institutions are supposed to ignore or reject payment instructions that don’t comport with precise formatting required by the National Automated Clearinghouse Association (NACHA), the not-for-profit organization that provides the backbone for the electronic movement of money in the United States. But Slavkin said a number of financial institutions ended up processing both reversal requests, meaning a fair number of employees at companies that use MyPayrollHR suddenly saw a month’s worth of payroll payments withdrawn from their bank accounts.

Dan L’Abbe, CEO of the San Francisco-based consultancy Granite Solutions Groupe, said the mix-up has been massively disruptive for his 250 employees.

“This caused a lot of chaos for employers, but employees were the ones really affected,” L’Abbe said. “This is all very unusual because we don’t even have the ability to take money out of our employee accounts.”

Slavkin said Cachet managed to reach the CEO of MyPayrollHR — Michael T. Mann — via phone on the evening of Sept. 4, and that Mann said he would would call back in a few minutes. According to Slavkin, Mann never returned the call. Not long after that, MyPayrollHR told clients that it was going out of business and that they should find someone else to handle their payroll.

In short order, many people hit by one or both payroll reversals took to Twitter and Facebook to vent their anger and bewilderment at Cachet and at MyPayrollHR. But Slavkin said Cachet ultimately decided to cancel the previous payment reversals, leaving Cachet on the hook for $26 million.

“What we have since done is reached out to 100+ receiving banks to have them reject both reversals,” Slavkin said. “So most — if not all — employees affected by this will in the next day or two have all their money back.”

THE VANISHING MANN

Cachet has since been in touch with the FBI and with federal prosecutors in New York, and Slavkin said both are now investigating MyPayrollHR and its CEO. On Monday, New York Governor Andrew Cuomo called on the state’s Department of Financial Services to investigate the company’s “sudden and disturbing shutdown.”

A tweet sent Sept. 11 by the FBI’s Albany field office.

The $26 million hit against Cachet wasn’t the only fraud apparently perpetrated by MyPayrollHR and/or its parent firm: According to Slavkin, the now defunct New York company also stiffed National Payment Corporation (NatPay) — the Florida-based firm which handles tax withholdings for MyPayrollHR clients — to the tune of more than $9 million.

In a statement provided to KrebsOnSecurity, NatPay said it was alerted late last week that the bank accounts of MyPayrollHR and one of its affiliated companies were frozen, and that the notification came after payment files were processed.

“NatPay was provided information that MyPayrollHR and Cloud Payroll may have been the victims of fraud committed by their holding company ValueWise, whose CEO and owner is Michael Mann,” NatPay said. “NatPay immediately put in place steps to manage the orderly process of recovering funds [and] has more than sufficient insurance to cover actions of attempted or real fraud.”

Requests for comment from different executives at both MyPayrollHR and its parent firm ValueWise Corp. went unanswered, and the latter’s Web site is now offline. Several erstwhile MyPayrollHR employees reached via LinkedIn said none of them had seen or heard from Mr. Mann in days.

Meanwhile, Granite Solutions Groupe CEO L’Abbe said some of his employees have seen their bank accounts credited back the money that was taken, while others are still waiting for those reversals to come through.

“It varies widely,” L’Abbe said. “Every bank processes differently, and everyone’s relationship with the bank is different. Others have absolutely no money right now and are having a helluva time with their bank believing this is all the result of fraud. Things are starting to settle down now, but a lot of employees are still in limbo with their bank.”

For its part, Cachet Financial says it will be looking at solutions to better detect when and if instructions from clients for funding its settlement accounts suddenly change.

“Our system is excellent at protecting against outside hackers,” Slavkin said. “But when it comes to something like this it takes everyone by complete surprise.”

Secret Service Investigates Breach at U.S. Govt IT Contractor

The U.S. Secret Service is investigating a breach at a Virginia-based government technology contractor that saw access to several of its systems put up for sale in the cybercrime underground, KrebsOnSecurity has learned. The contractor claims the access being auctioned off was to old test systems that do not have direct connections to its government partner networks.

In mid-August, a member of a popular Russian-language cybercrime forum offered to sell access to the internal network of a U.S. government IT contractor that does business with more than 20 federal agencies, including several branches of the military. The seller bragged that he had access to email correspondence and credentials needed to view databases of the client agencies, and set the opening price at six bitcoins (~USD $60,000).

A review of the screenshots posted to the cybercrime forum as evidence of the unauthorized access revealed several Internet addresses tied to systems at the U.S. Department of Transportation, the National Institutes of Health (NIH), and U.S. Citizenship and Immigration Services (USCIS), a component of the U.S. Department of Homeland Security that manages the nation’s naturalization and immigration system.

Other domains and Internet addresses included in those screenshots pointed to Miracle Systems LLC, an Arlington, Va. based IT contractor that states on its site that it serves 20+ federal agencies as a prime contractor, including the aforementioned agencies.

In an interview with KrebsOnSecurity, Miracle Systems CEO Sandesh Sharda confirmed that the auction concerned credentials and databases managed by his company, and that an investigating agent from the Secret Service was in his firm’s offices at that very moment looking into the matter.

But he maintained that the purloined data shown in the screenshots was years-old and mapped only to internal test systems that were never connected to its government agency clients.

“The Secret Service came to us and said they’re looking into the issue,” Sharda said. “But it was all old stuff [that was] in our own internal test environment, and it is no longer valid.”

Still, Sharda did acknowledge information shared by Wisconsin-based security firm Hold Security, which alerted KrebsOnSecurity to this incident, indicating that at least eight of its internal systems had been compromised on three separate occasions between November 2018 and July 2019 by Emotet, a malware strain usually distributed via malware-laced email attachments that typically is used to deploy other malicious software.

The Department of Homeland Security did not respond to requests for comment, nor did the Department of Transportation. A spokesperson for the NIH said the agency had investigated the activity and found it was not compromised by the incident.

“As is the case for all agencies of the Federal Government, the NIH is constantly under threat of cyber-attack,” NIH spokesperson Julius Patterson said. “The NIH has a comprehensive security program that is continuously monitoring and responding to security events, and cyber-related incidents are reported to the Department of Homeland Security through the HHS Computer Security Incident Response Center.”

One of several screenshots offered by the dark web seller as proof of access to a federal IT contractor later identified as Arlington, Va. based Miracle Systems. Image: Hold Security.

The dust-up involving Miracle Systems comes amid much hand-wringing among U.S. federal agencies about how best to beef up and ensure security at a slew of private companies that manage federal IT contracts and handle government data.

For years, federal agencies had few options to hold private contractors to the same security standards to which they must adhere — beyond perhaps restricting how federal dollars are spent. But recent updates to federal acquisition regulations allow agencies to extend those same rules to vendors, enforce specific security requirements, and even kill contracts that are found to be in violation of specific security clauses.

In July, DHS’s Customs and Border Patrol (CPB) suspended all federal contracts with Perceptics, a contractor which sells license-plate scanners and other border control equipment, after data collected by the company was made available for download on the dark web. The CPB later said the breach was the result of a federal contractor copying data on its corporate network, which was subsequently compromised.

For its part, the Department of Defense recently issued long-awaited cybersecurity standards for contractors who work with the Pentagon’s sensitive data.

“This problem is not necessarily a tier-one supply level,” DOD Chief Information Officer Dana Deasy told the Senate Armed Services Committee earlier this year. “It’s down when you get to the tier-three and the tier-four” subcontractors.

Spam In your Calendar? Here’s What to Do.

Many spam trends are cyclical: Spammers tend to switch tactics when one method of hijacking your time and attention stops working. But periodically they circle back to old tricks, and few spam trends are as perennial as calendar spam, in which invitations to click on dodgy links show up unbidden in your digital calendar application from Apple, Google and Microsoft. Here’s a brief primer on what you can do about it.

Image: Reddit

Over the past few weeks, a good number of readers have written in to say they feared their calendar app or email account was hacked after noticing a spammy event had been added to their calendars.

The truth is, all that a spammer needs to add an unwelcome appointment to your calendar is the email address tied to your calendar account. That’s because the calendar applications from Apple, Google and Microsoft are set by default to accept calendar invites from anyone.

Calendar invites from spammers run the gamut from ads for porn or pharmacy sites, to claims of an unexpected financial windfall or “free” items of value, to outright phishing attacks and malware lures. The important thing is that you don’t click on any links embedded in these appointments. And resist the temptation to respond to such invitations by selecting “yes,” “no,” or “maybe,” as doing so may only serve to guarantee you more calendar spam.

Fortunately, the are a few simple steps you can take that should help minimize this nuisance. To stop events from being automatically added to your Google calendar:

-Open the Calendar application, and click the gear icon to get to the Calendar Settings page.
-Under “Event Settings,” change the default setting to “No, only show invitations to which I have responded.”

To prevent events from automatically being added to your Microsoft Outlook calendar, click the gear icon in the upper right corner of Outlook to open the settings menu, and then scroll down and select “View all Outlook settings.” From there:

-Click “Calendar,” then “Events from email.”
-Change the default setting for each type of reservation settings to “Only show event summaries in email.”

For Apple calendar users, log in to your iCloud.com account, and select Calendar.

-Click the gear icon in the lower left corner of the Calendar application, and select “Preferences.”
-Click the “Advanced” tab at the top of the box that appears.
-Change the default setting to “Email to [your email here].”

Making these changes will mean that any events your email provider previously added to your calendar automatically by scanning your inbox for certain types of messages from common events — such as making hotel, dining, plane or train reservations, or paying recurring bills — may no longer be added for you. Spammy calendar invitations may still show up via email; in the event they do, make sure to mark the missives as spam.

Have you experienced a spike in calendar spam of late? Or maybe you have another suggestion for blocking it? If so, sound off in the comments below.

‘Satori’ IoT Botnet Operator Pleads Guilty

A 21-year-old man from Vancouver, Wash. has pleaded guilty to federal hacking charges tied to his role in operating the “Satori” botnet, a crime machine powered by hacked Internet of Things (IoT) devices that was built to conduct massive denial-of-service attacks targeting Internet service providers, online gaming platforms and Web hosting companies.

Kenneth “Nexus-Zeta” Schuchman, in an undated photo.

Kenneth Currin Schuchman pleaded guilty to one count of aiding and abetting computer intrusions. Between July 2017 and October 2018, Schuchman was part of a conspiracy with at least two other unnamed individuals to develop and use Satori in large scale online attacks designed to flood their targets with so much junk Internet traffic that the targets became unreachable by legitimate visitors.

According to his plea agreement, Schuchman — who went by the online aliases “Nexus” and “Nexus-Zeta” — worked with at least two other individuals to build and use the Satori botnet, which harnessed the collective bandwidth of approximately 100,000 hacked IoT devices by exploiting vulnerabilities in various wireless routers, digital video recorders, Internet-connected security cameras, and fiber-optic networking devices.

Satori was originally based on the leaked source code for Mirai, a powerful IoT botnet that first appeared in the summer of 2016 and was responsible for some of the largest denial-of-service attacks ever recorded (including a 620 Gbps attack that took KrebsOnSecurity offline for almost four days).

Throughout 2017 and into 2018, Schuchman worked with his co-conspirators — who used the nicknames “Vamp” and “Drake” — to further develop Satori by identifying and exploiting additional security flaws in other IoT systems.

Schuchman and his accomplices gave new monikers to their IoT botnets with almost each new improvement, rechristening their creations with names including “Okiru,” and “Masuta,” and infecting up to 700,000 compromised systems.

The plea agreement states that the object of the conspiracy was to sell access to their botnets to those who wished to rent them for launching attacks against others, although it’s not clear to what extent Schuchman and his alleged co-conspirators succeeded in this regard.

Even after he was indicted in connection with his activities in August 2018, Schuchman created a new botnet variant while on supervised release. At the time, Schuchman and Drake had something of a falling out, and Schuchman later acknowledged using information gleaned by prosecutors to identify Drake’s home address for the purposes of “swatting” him.

Swatting involves making false reports of a potentially violent incident — usually a phony hostage situation, bomb threat or murder — to prompt a heavily-armed police response to the target’s location. According to his plea agreement, the swatting that Schuchman set in motion in October 2018 resulted in “a substantial law enforcement response at Drake’s residence.”

As noted in a September 2018 story, Schuchman was not exactly skilled in the art of obscuring his real identity online. For one thing, the domain name used as a control server to synchronize the activities of the Satori botnet was registered to the email address nexuczeta1337@gmail.com. That domain name was originally registered to a “ZetaSec Inc.” and to a “Kenny Schuchman” in Vancouver, Wash.

People who operate IoT-based botnets maintain and build up their pool of infected IoT systems by constantly scanning the Internet for other vulnerable systems. Schuchman’s plea agreement states that when he received abuse complaints related to his scanning activities, he responded in his father’s identity.

“Schuchman frequently used identification devices belonging to his father to further the criminal scheme,” the plea agreement explains.

While Schuchman may be the first person to plead guilty in connection with Satori and its progeny, he appears to be hardly the most culpable. Multiple sources tell KrebsOnSecurity that Schuchman’s co-conspirator Vamp is a U.K. resident who was principally responsible for coding the Satori botnet, and as a minor was involved in the 2015 hack against U.K. phone and broadband provider TalkTalk.

Multiple sources also say Vamp was principally responsible for the 2016 massive denial-of-service attack that swamped Dyn — a company that provides core Internet services for a host of big-name Web sites. On October 21, 2016, an attack by a Mirai-based IoT botnet variant overwhelmed Dyn’s infrastructure, causing outages at a number of top Internet destinations, including Twitter, Spotify, Reddit and others.

The investigation into Schuchman and his alleged co-conspirators is being run out the FBI field office in Alaska, spearheaded by some of the same agents who helped track down and ultimately secure guilty pleas from the original co-authors of the Mirai botnet.

It remains to be seen what kind of punishment a federal judge will hand down for Schuchman, who reportedly has been diagnosed with Asperger Syndrome and autism. The maximum penalty for the single criminal count to which he’s pleaded guilty is 10 years in prison and fines of up to $250,000.

However, it seems likely his sentencing will fall well short of that maximum: Schuchman’s plea deal states that he agreed to a recommended sentence “at the low end of the guideline range as calculated and adopted by the court.”

Feds Allege Adconion Employees Hijacked IP Addresses for Spamming

Federal prosecutors in California have filed criminal charges against four employees of Adconion Direct, an email advertising firm, alleging they unlawfully hijacked vast swaths of Internet addresses and used them in large-scale spam campaigns. KrebsOnSecurity has learned that the charges are likely just the opening salvo in a much larger, ongoing federal investigation into the company’s commercial email practices.

Prior to its acquisition, Adconion offered digital advertising solutions to some of the world’s biggest companies, including Adidas, AT&T, Fidelity, Honda, Kohl’s and T-Mobile. Amobee, the Redwood City, Calif. online ad firm that acquired Adconion in 2014, bills itself as the world’s leading independent advertising platform. The CEO of Amobee is Kim Perell, formerly CEO of Adconion.

In October 2018, prosecutors in the Southern District of California named four Adconion employees — Jacob Bychak, Mark ManoogianPetr Pacas, and Mohammed Abdul Qayyum —  in a ten-count indictment on charges of conspiracy, wire fraud, and electronic mail fraud. All four men have pleaded not guilty to the charges, which stem from a grand jury indictment handed down in June 2017.

‘COMPANY A’

The indictment and other court filings in this case refer to the employer of the four men only as “Company A.” However, LinkedIn profiles under the names of three of the accused show they each work(ed) for Adconion and/or Amobee.

Mark Manoogian is an attorney whose LinkedIn profile states that he is director of legal and business affairs at Amobee, and formerly was senior business development manager at Adconion Direct; Bychak is listed as director of operations at Adconion Direct; Quayyum’s LinkedIn page lists him as manager of technical operations at Adconion. A statement of facts filed by the government indicates Petr Pacas was at one point director of operations at Company A (Adconion).

According to the indictment, between December 2010 and September 2014 the defendants engaged in a conspiracy to identify or pay to identify blocks of Internet Protocol (IP) addresses that were registered to others but which were otherwise inactive.

The government alleges the men sent forged letters to an Internet hosting firm claiming they had been authorized by the registrants of the inactive IP addresses to use that space for their own purposes.

“Members of the conspiracy would use the fraudulently acquired IP addresses to send commercial email (‘spam’) messages,” the government charged.

HOSTING IN THE WIND

Prosecutors say the accused were able to spam from the purloined IP address blocks after tricking the owner of Hostwinds, an Oklahoma-based Internet hosting firm, into routing the fraudulently obtained IP addresses on their behalf.

Hostwinds owner Peter Holden was the subject of a 2015 KrebsOnSecurity story titled, “Like Cutting Off a Limb to Save the Body,” which described how he’d initially built a lucrative business catering mainly to spammers, only to later have a change of heart and aggressively work to keep spammers off of his network.

That a case of such potential import for the digital marketing industry has escaped any media attention for so long is unusual but not surprising given what’s at stake for the companies involved and for the government’s ongoing investigations.

Adconion’s parent Amobee manages ad campaigns for some of the world’s top brands, and has every reason not to call attention to charges that some of its key employees may have been involved in criminal activity.

Meanwhile, prosecutors are busy following up on evidence supplied by several cooperating witnesses in this and a related grand jury investigation, including a confidential informant who received information from an Adconion employee about the company’s internal operations.

THE BIGGER PICTURE

According to a memo jointly filed by the defendants, “this case spun off from a larger ongoing investigation into the commercial email practices of Company A.” Ironically, this memo appears to be the only one of several dozen documents related to the indictment that mentions Adconion by name (albeit only in a series of footnote references).

Prosecutors allege the four men bought hijacked IP address blocks from another man tied to this case who was charged separately. This individual, Daniel Dye, has a history of working with others to hijack IP addresses for use by spammers.

For many years, Dye was a system administrator for Optinrealbig, a Colorado company that relentlessly pimped all manner of junk email, from mortgage leads and adult-related services to counterfeit products and Viagra.

Optinrealbig’s CEO was the spam king Scott Richter, who later changed the name of the company to Media Breakaway after being successfully sued for spamming by AOL, MicrosoftMySpace, and the New York Attorney General Office, among others. In 2008, this author penned a column for The Washington Post detailing how Media Breakaway had hijacked tens of thousands of IP addresses from a defunct San Francisco company for use in its spamming operations.

Dye has been charged with violations of the CAN-SPAM Act. A review of the documents in his case suggest Dye accepted a guilty plea agreement in connection with the IP address thefts and is cooperating with the government’s ongoing investigation into Adconion’s email marketing practices, although the plea agreement itself remains under seal.

Lawyers for the four defendants in this case have asserted in court filings that the government’s confidential informant is an employee of Spamhaus.org, an organization that many Internet service providers around the world rely upon to help identify and block sources of malware and spam.

Interestingly, in 2014 Spamhaus was sued by Blackstar Media LLC, a bulk email marketing company and subsidiary of Adconion. Blackstar’s owners sued Spamhaus for defamation after Spamhaus included them at the top of its list of the Top 10 world’s worst spammers. Blackstar later dropped the lawsuit and agreed to paid Spamhaus’ legal costs.

Representatives for Spamhaus declined to comment for this story. Responding to questions about the indictment of Adconion employees, Amobee’s parent company SingTel referred comments to Amobee, which issued a brief statement saying, “Amobee has fully cooperated with the government’s investigation of this 2017 matter which pertains to alleged activities that occurred years prior to Amobee’s acquisition of the company.”

ONE OF THE LARGEST SPAMMERS IN HISTORY?

It appears the government has been investigating Adconion’s email practices since at least 2015, and possibly as early as 2013. The very first result in an online search for the words “Adconion” and “spam” returns a Microsoft Powerpoint document that was presented alongside this talk at an ARIN meeting in October 2016. ARIN stands for the American Registry for Internet Numbers, and it handles IP addresses allocations for entities in the United States, Canada and parts of the Caribbean.

As the screenshot above shows, that Powerpoint deck was originally named “Adconion – Arin,” but the file has since been renamed. That is, unless one downloads the file and looks at the metadata attached to it, which shows the original filename and that it was created in 2015 by someone at the U.S. Department of Justice.

Slide #8 in that Powerpoint document references a case example of an unnamed company (again, “Company A”), which the presenter said was “alleged to be one of the largest spammers in history,” that had hijacked “hundreds of thousands of IP addresses.”

A slide from an ARIN presentation in 2016 that referenced Adconion.

There are fewer than four billion IPv4 addresses available for use, but the vast majority of them have already been allocated. In recent years, this global shortage has turned IP addresses into a commodity wherein each IP can fetch between $15-$25 on the open market.

The dearth of available IP addresses has created boom times for those engaged in the acquisition and sale of IP address blocks. It also has emboldened scammers and spammers who specialize in absconding with and spamming from dormant IP address blocks without permission from the rightful owners.

In May, KrebsOnSecurity broke the news that Amir Golestan — the owner of a prominent Charleston, S.C. tech company called Micfo LLC — had been indicted on criminal charges of fraudulently obtaining more than 735,000 IP addresses from ARIN and reselling the space to others.

KrebsOnSecurity has since learned that for several years prior to 2014, Adconion was one of Golestan’s biggest clients. More on that in an upcoming story.

Ransomware Bites Dental Data Backup Firm

PerCSoft, a Wisconsin-based company that manages a remote data backup service relied upon by hundreds of dental offices across the country, is struggling to restore access to client systems after falling victim to a ransomware attack.

West Allis, Wis.-based PerCSoft is a cloud management provider for Digital Dental Record (DDR), which operates an online data backup service called DDS Safe that archives medical records, charts, insurance documents and other personal information for various dental offices across the United States.

The ransomware attack hit PerCSoft on the morning of Monday, Aug. 26, and encrypted dental records for some — but not all — of the practices that rely on DDS Safe.

PercSoft did not respond to requests for comment. But Brenna Sadler, director of  communications for the Wisconsin Dental Association, said the ransomware encrypted files for approximate 400 dental practices, and that somewhere between 80-100 of those clients have now had their files restored.

Sadler said she did not know whether PerCSoft and/or DDR had paid the ransom demand, what ransomware strain was involved, or how much the attackers had demanded.

But updates to PerCSoft’s Facebook page and statements published by both PerCSoft and DDR suggest someone may have paid up: The statements note that both companies worked with a third party software company and were able to obtain a decryptor to help clients regain access to files that were locked by the ransomware.

Update: Several sources are now reporting that PerCSoft did pay the ransom, although it is not clear how much was paid. One member of a private Facebook group dedicated to IT professionals serving the dental industry shared the following screenshot, which is purportedly from a conversation between PerCSoft and an affected dental office, indicating the cloud provider was planning to pay the ransom:

Another image shared by members of that Facebook group indicates the ransomware that attacked PerCSoft is an extremely advanced and fairly recent strain known variously as REvil and Sodinokibi.

Original story:

However, some affected dental offices have reported that the decryptor did not work to unlock at least some of the files encrypted by the ransomware. Meanwhile, several affected dentistry practices said they feared they might be unable to process payroll payments this week as a result of the attack.

Cloud data and backup services are a prime target of cybercriminals who deploy ransomware. In July, attackers hit QuickBooks cloud hosting firm iNSYNQ, holding data hostage for many of the company’s clients. In February, cloud payroll data provider Apex Human Capital Management was knocked offline for three days following a ransomware infestation.

On Christmas Eve 2018, cloud hosting provider Dataresolution.net took its systems offline in response to a ransomware outbreak on its internal networks. The company was adamant that it would not pay the ransom demand, but it ended up taking several weeks for customers to fully regain access to their data.

The FBI and multiple security firms have advised victims not to pay any ransom demands, as doing so just encourages the attackers and in any case may not result in actually regaining access to encrypted files. In practice, however, many cybersecurity consulting firms are quietly urging their customers that paying up is the fastest route back to business-as-usual.

It remains unclear whether PerCSoft or DDR — or perhaps their insurance provider — paid the ransom demand in this attack. But new reporting from independent news outlet ProPublica this week sheds light on another possible explanation why so many victims are simply coughing up the money: Their insurance providers will cover the cost — minus a deductible that is usually far less than the total ransom demanded by the attackers.

More to the point, ProPublica found, such attacks may be great for business if you’re in the insurance industry.

“More often than not, paying the ransom is a lot cheaper for insurers than the loss of revenue they have to cover otherwise,” said Minhee Cho, public relations director of ProPublica, in an email to KrebsOnSecurity. “But, by rewarding hackers, these companies have created a perverted cycle that encourages more ransomware attacks, which in turn frighten more businesses and government agencies into buying policies.”

“In fact, it seems hackers are specifically extorting American companies that they know have cyber insurance,” Cho continued. “After one small insurer highlighted the names of some of its cyber policyholders on its website, three of them were attacked by ransomware.”

Read the full ProPublica piece here. And if you haven’t already done so, check out this outstanding related reporting by ProPublica from earlier this year on how security firms that help companies respond to ransomware attacks also may be enabling and emboldening attackers.

Phishers are Angling for Your Cloud Providers

Many companies are now outsourcing their marketing efforts to cloud-based Customer Relationship Management (CRM) providers. But when accounts at those CRM providers get hacked or phished, the results can be damaging for both the client’s brand and their customers. Here’s a look at a recent CRM-based phishing campaign that targeted customers of Fortune 500 construction equipment vendor United Rentals.

Stamford, Ct.-based United Rentals [NYSE:URI] is the world’s largest equipment rental company, with some 18,000 employees and earnings of approximately $4 billion in 2018. On August 21, multiple United Rental customers reported receiving invoice emails with booby-trapped links that led to a malware download for anyone who clicked.

While phony invoices are a common malware lure, this particular campaign sent users to a page on United Rentals’ own Web site (unitedrentals.com).

A screen shot of the malicious email that spoofed United Rentals.

In a notice to customers, the company said the unauthorized messages were not sent by United Rentals. One source who had at least two employees fall for the scheme forwarded KrebsOnSecurity a response from UR’s privacy division, which blamed the incident on a third-party advertising partner.

“Based on current knowledge, we believe that an unauthorized party gained access to a vendor platform United Rentals uses in connection with designing and executing email campaigns,” the response read.

“The unauthorized party was able to send a phishing email that appears to be from United Rentals through this platform,” the reply continued. “The phishing email contained links to a purported invoice that, if clicked on, could deliver malware to the recipient’s system. While our investigation is continuing, we currently have no reason to believe that there was unauthorized access to the United Rentals systems used by customers, or to any internal United Rentals systems.”

United Rentals told KrebsOnSecurity that its investigation so far reveals no compromise of its internal systems.

“At this point, we believe this to be an email phishing incident in which an unauthorized third party used a third-party system to generate an email campaign to deliver what we believe to be a banking trojan,” said Dan Higgins, UR’s chief information officer.

United Rentals would not name the third party marketing firm thought to be involved, but passive DNS lookups on the UR subdomain referenced in the phishing email (used by UL for marketing since 2014 and visible in the screenshot above as “wVw.unitedrentals.com”) points to Pardot, an email marketing division of cloud CRM giant Salesforce.

Companies that use cloud-based CRMs sometimes will dedicate a domain or subdomain they own specifically for use by their CRM provider, allowing the CRM to send emails that appear to come directly from the client’s own domains. However, in such setups the content that gets promoted through the client’s domain is actually hosted on the cloud CRM provider’s systems.

Salesforce did not respond to multiple requests for comment. But it seems likely that someone at Pardot with access to United Rental’s account was phished, hacked, or perhaps guilty of password re-use.

This attack comes on the heels of another targeted phishing campaign leveraging Pardot that was documented earlier this month by Netskope, a cloud security firm. Netskope’s Ashwin Vamshi said users of cloud CRM platforms have a high level of trust in the software because they view the data and associated links as internal, even though they are hosted in the cloud.

“A large number of enterprises provide their vendors and partners access to their CRM for uploading documents such as invoices, purchase orders, etc. (and often these happen as automated workflows),” Vamshi wrote. “The enterprise has no control over the vendor or partner device and, more importantly, over the files being uploaded from them. In many cases, vendor- or partner-uploaded files carry with them a high level of implicit trust.”

Cybercriminals increasingly are targeting cloud CRM providers because compromised accounts on these systems can be leveraged to conduct extremely targeted and convincing phishing attacks. According to the most recent stats (PDF) from the Anti-Phishing Working Group, software-as-a-service providers (including CRM and Webmail providers) were the most-targeted industry sector in the first quarter of 2019, accounting for 36 percent of all phishing attacks.

Image: APWG

Cybersecurity Firm Imperva Discloses Breach

Imperva, a leading provider of Internet firewall services that help Web sites block malicious cyberattacks, alerted customers on Tuesday that a recent data breach exposed email addresses, scrambled passwords, API keys and SSL certificates for a subset of its firewall users.

Redwood Shores, Calif.-based Imperva sells technology and services designed to detect and block various types of malicious Web traffic, from denial-of-service attacks to digital probes aimed at undermining the security of Web-based software applications.

Image: Imperva

Earlier today, Imperva told customers that it learned on Aug. 20 about a security incident that exposed sensitive information for some users of Incapsula, the company’s cloud-based Web Application Firewall (WAF) product.

“On August 20, 2019, we learned from a third party of a data exposure that impacts a subset of customers of our Cloud WAF product who had accounts through September 15, 2017,” wrote Heli Erickson, director of analyst relations at Imperva.

“We want to be very clear that this data exposure is limited to our Cloud WAF product,” Erickson’s message continued. “While the situation remains under investigation, what we know today is that elements of our Incapsula customer database from 2017, including email addresses and hashed and salted passwords, and, for a subset of the Incapsula customers from 2017, API keys and customer-provided SSL certificates, were exposed.”

Companies that use the Incapsula WAF route all of their Web site traffic through the service, which scrubs the communications for any suspicious activity or attacks and then forwards the benign traffic on to its intended destination.

Rich Mogull, founder and vice president of product at Kansas City-based cloud security firm DisruptOps, said Imperva is among the top three Web-based firewall providers in business today.

According to Mogull, an attacker in possession of a customer’s API keys and SSL certificates could use that access to significantly undermine the security of traffic flowing to and from a customer’s various Web sites.

At a minimum, he said, an attacker in possession of these key assets could reduce the security of the WAF settings and exempt or “whitelist” from the WAF’s scrubbing technology any traffic coming from the attacker. A worst-case scenario could allow an attacker to intercept, view or modify traffic destined for an Incapsula client Web site, and even to divert all traffic for that site to or through a site owned by the attacker.

“Attackers could whitelist themselves and begin attacking the site without the WAF’s protection,” Mogull told KrebsOnSecurity. “They could modify any of the security Incapsula security settings, and if they got [the target’s SSL] certificate, that can potentially expose traffic. For a security-as-a-service provider like Imperva, this is the kind of mistake that’s up their with their worst nightmare.”

Imperva urged all of its customers to take several steps that might mitigate the threat from the data exposure, such as changing passwords for user accounts at Incapsula, enabling multi-factor authentication, resetting API keys, and generating/uploading new SSL certificates.

Alissa Knight, a senior analyst at Aite Group, said the exposure of Incapsula users’ scrambled passwords and email addresses was almost incidental given that the intruders also made off with customer API keys and SSL certificates.

Knight said although we don’t yet know the cause of this incident, such breaches at cloud-based firms often come down to small but ultimately significant security failures on the part of the provider.

“The moral of the story here is that people need to be asking tough questions of software-as-a-service firms they rely upon, because those vendors are being trusted with the keys to the kingdom,” Knight said. “Even if the vendor in question is a cybersecurity company, it doesn’t necessarily mean they’re eating their own dog food.”

Breach at Hy-Vee Supermarket Chain Tied to Sale of 5M+ Stolen Credit, Debit Cards

On Tuesday of this week, one of the more popular underground stores peddling credit and debit card data stolen from hacked merchants announced a blockbuster new sale: More than 5.3 million new accounts belonging to cardholders from 35 U.S. states. Multiple sources now tell KrebsOnSecurity that the card data came from compromised gas pumps, coffee shops and restaurants operated by Hy-Vee, an Iowa-based company that operates a chain of more than 245 supermarkets throughout the Midwestern United States.

Hy-Vee, based in Des Moines, announced on Aug. 14 it was investigating a data breach involving payment processing systems that handle transactions at some Hy-Vee fuel pumps, drive-thru coffee shops and restaurants.

The restaurants affected include Hy-Vee Market Grilles, Market Grille Expresses and Wahlburgers locations that the company owns and operates. Hy-Vee said it was too early to tell when the breach initially began or for how long intruders were inside their payment systems.

But typically, such breaches occur when cybercriminals manage to remotely install malicious software on a retailer’s card-processing systems. This type of point-of-sale malware is capable of copying data stored on a credit or debit card’s magnetic stripe when those cards are swiped at compromised payment terminals. This data can then be used to create counterfeit copies of the cards.

Hy-Vee said it believes the breach does not affect payment card terminals used at its grocery store checkout lanes, pharmacies or convenience stores, as these systems rely on a security technology designed to defeat card-skimming malware.

“These locations have different point-of-sale systems than those located at our grocery stores, drugstores and inside our convenience stores, which utilize point-to-point encryption technology for processing payment card transactions,” Hy-Vee said. “This encryption technology protects card data by making it unreadable. Based on our preliminary investigation, we believe payment card transactions that were swiped or inserted on these systems, which are utilized at our front-end checkout lanes, pharmacies, customer service counters, wine & spirits locations, floral departments, clinics and all other food service areas, as well as transactions processed through Aisles Online, are not involved.”

According to two sources who asked not to be identified for this story — including one at a major U.S. financial institution — the card data stolen from Hy-Vee is now being sold under the code name “Solar Energy,” at the infamous Joker’s Stash carding bazaar.

An ad at the Joker’s Stash carding site for “Solar Energy,” a batch of more than 5 million credit and debit cards sources say was stolen from customers of supermarket chain Hy-Vee.

Hy-Vee said the company’s investigation is continuing.

“We are aware of reports from payment processors and the card networks of payment data being offered for sale and are working with the payment card networks so that they can identify the cards and work with issuing banks to initiate heightened monitoring on accounts,” Hy-Vee spokesperson Tina Pothoff said.

The card account records sold by Joker’s Stash, known as “dumps,” apparently stolen from Hy-Vee are being sold for prices ranging from $17 to $35 apiece. Buyers typically receive a text file that includes all of their dumps. Those individual dumps records — when encoded onto a new magnetic stripe on virtually anything the size of a credit card — can be used to purchase stolen merchandise in big box stores.

As noted in previous stories here, the organized cyberthieves involved in stealing card data from main street merchants have gradually moved down the food chain from big box retailers like Target and Home Depot to smaller but far more plentiful and probably less secure merchants (either by choice or because the larger stores became a harder target).

It’s really not worth spending time worrying about where your card number may have been breached, since it’s almost always impossible to say for sure and because it’s common for the same card to be breached at multiple establishments during the same time period.

Just remember that while consumers are not liable for fraudulent charges, it may still fall to you the consumer to spot and report any suspicious charges. So keep a close eye on your statements, and consider signing up for text message notifications of new charges if your card issuer offers this service. Most of these services also can be set to alert you if you’re about to miss an upcoming payment, so they can also be handy for avoiding late fees and other costly charges.

Forced Password Reset? Check Your Assumptions

Almost weekly now I hear from an indignant reader who suspects a data breach at a Web site they frequent that has just asked the reader to reset their password. Further investigation almost invariably reveals that the password reset demand was not the result of a breach but rather the site’s efforts to identify customers who are reusing passwords from other sites that have already been hacked.

But ironically, many companies taking these proactive steps soon discover that their explanation as to why they’re doing it can get misinterpreted as more evidence of lax security. This post attempts to unravel what’s going on here.

Over the weekend, a follower on Twitter included me in a tweet sent to California-based job search site Glassdoor, which had just sent him the following notice:

The Twitter follower expressed concern about this message, because it suggested to him that in order for Glassdoor to have done what it described, the company would have had to be storing its users’ passwords in plain text. I replied that this was in fact not an indication of storing passwords in plain text, and that many companies are now testing their users’ credentials against lists of hacked credentials that have been leaked and made available online.

The reality is Facebook, Netflix and a number of big-name companies are regularly combing through huge data leak troves for credentials that match those of their customers, and then forcing a password reset for those users. Some are even checking for password re-use on all new account signups.

The idea here is to stymie a massively pervasive problem facing all companies that do business online today: Namely, “credential-stuffing attacks,” in which attackers take millions or even billions of email addresses and corresponding cracked passwords from compromised databases and see how many of them work at other online properties.

So how does the defense against this daily deluge of credential stuffing work? A company employing this strategy will first extract from these leaked credential lists any email addresses that correspond to their current user base.

From there, the corresponding cracked (plain text) passwords are fed into the same process that the company relies upon when users log in: That is, the company feeds those plain text passwords through its own password “hashing” or scrambling routine.

Password hashing is designed to be a one-way function which scrambles a plain text password so that it produces a long string of numbers and letters. Not all hashing methods are created equal, and some of the most commonly used methods — MD5 and SHA-1, for example — can be far less secure than others, depending on how they’re implemented (more on that in a moment). Whatever the hashing method used, it’s the hashed output that gets stored, not the password itself.

Back to the process: If a user’s plain text password from a hacked database matches the output of what a company would expect to see after running it through their own internal hashing process, that user is then prompted to change their password to something truly unique.

Now, password hashing methods can be made more secure by amending the password with what’s known as a “salt” — or random data added to the input of a hash function to guarantee a unique output. And many readers of the Twitter thread on Glassdoor’s approach reasoned that the company couldn’t have been doing what it described without also forgoing this additional layer of security.

My tweeted explanatory reply as to why Glassdoor was doing this was (in hindsight) incomplete and in any case not as clear as it should have been. Fortunately, Glassdoor’s chief information officer Anthony Moisant chimed in to the Twitter thread to explain that the salt is in fact added as part of the password testing procedure.

“In our [user] database, we’ve got three columns — username, salt value and scrypt hash,” Moisant explained in an interview with KrebsOnSecurity. “We apply the salt that’s stored in the database and the hash [function] to the plain text password, and that resulting value is then checked against the hash in the database we store. For whatever reason, some people have gotten it into their heads that there’s no possible way to do these checks if you salt, but that’s not true.”

CHECK YOUR ASSUMPTIONS

You — the user — can’t be expected to know or control what password hashing methods a given site uses, if indeed they use them at all. But you can control the quality of the passwords you pick.

I can’t stress this enough: Do not re-use passwords. And don’t recycle them either. Recycling involves rather lame attempts to make a reused password unique by simply adding a digit or changing the capitalization of certain characters. Crooks who specialize in password attacks are wise to this approach as well.

If you have trouble remembering complex passwords (and this describes most people), consider relying instead on password length, which is a far more important determiner of whether a given password can be cracked by available tools in any timeframe that might be reasonably useful to an attacker.

In that vein, it’s safer and wiser to focus on picking passphrases instead of passwords. Passphrases are collections of multiple (ideally unrelated) words mushed together. Passphrases are not only generally more secure, they also have the added benefit of being easier to remember.

According to a recent blog entry by Microsoft group program manager Alex Weinert, none of the above advice about password complexity amounts to a hill of beans from the attacker’s standpoint.

Weinert’s post makes a compelling argument that as long as we’re stuck with passwords, taking full advantage of the most robust form of multi-factor authentication (MFA) offered by a site you frequent is the best way to deter attackers. Twofactorauth.org has a handy list of your options here, broken down by industry.

“Your password doesn’t matter, but MFA does,” Weinert wrote. “Based on our studies, your account is more than 99.9% less likely to be compromised if you use MFA.”

Glassdoor’s Moisant said the company doesn’t currently offer MFA for its users, but that it is planning to roll that out later this year to both consumer and business users.

Password managers also can be useful for those who feel encumbered by having to come up with passphrases or complex passwords. If you’re uncomfortable with entrusting a third-party service or application to handle this process for you, there’s absolutely nothing wrong with writing down your passwords, provided a) you do not store them in a file on your computer or taped to your laptop or screen or whatever, and b) that your password notebook is stored somewhere relatively secure, i.e. not in your purse or car, but something like a locked drawer or safe.

Although many readers will no doubt take me to task on that last bit of advice, as in all things security related it’s important not to let the perfect become the enemy of the good. Many people (think moms/dads/grandparents) can’t be bothered to use password managers  — even when you go through the trouble of setting them up on their behalf. Instead, without an easier, non-technical method they will simply revert to reusing or recycling passwords.