Topics:

Is Visa Making Up Compensation, Fine Calculations? Court Filings Raise Questions

Tools

No retailer likes being fined by Visa or MasterCard for letting thieves steal payment-card data, and most grumble privately about how that process is arbitrary and rigged against merchants. But a lawsuit now unfolding in Utah between a processor and a small restaurant has uncovered a remarkable level of detail about how arbitrary card brands can be—and with a Washington, D.C., lobbyist now backing the restaurant, it may also represent a real challenge to PCI fines.

The lawsuit is challenging everything from issuing banks' contracts to Visa's claims for counting up card fraud and pinpointing who's to blame—in addition to $1.3 million in card fraud that Visa says the restaurant enabled via an alleged security breach for which there's no concrete evidence.

By their nature, lawsuits are he-said/she-said affairs—that's why they go to trial, which could be years away in this case. But what all the parties seem to agree on is that in March 2008, Visa notified acquirer U.S. Bank of a potential data breach at one of its customers, a restaurant named Cisero's in the Utah ski resort town of Park City.

Over the next few months, Visa calculated that more $1.3 million in "actual fraud" was performed with card numbers stolen from Cisero's, and then fined U.S. Bank $80,000, which the bank and its processor, Elavon, passed on to Cisero's. MasterCard fined U.S. Bank $15,000, which it passed on, too—although only about $10,000 was removed from Cisero's merchant account before the restaurant changed banks.

Meanwhile, Cisero's went through two internal audits and two forensic investigations, which "revealed no concrete evidence that the POS server suffered a security breach" and "revealed no evidence of intrusive, malicious or unauthorized activity" on the hard drives of Cisero's servers, according to a forensic investigation by Cybertrust, a company certified by Visa and MasterCard. A forensic investigation by another company, Cadence Assurance, came to the same conclusion.

Visa's determination that fraudulently used card numbers came from Cisero's appears to have been solely from Visa's common point of purchase analysis. Cisero's claims that Visa has never explained or documented the $1.3 million in actual loss that Visa said it identified.

But Cybertrust did find payment-card numbers on a Cisero's hard drive, including 8,107 different Visa card numbers used in a total of 22,700 transactions. That's a PCI violation, even though Cisero's had been assured three years earlier that there were no card numbers stored on the drive when its POS vendor installed new software that was identified as PCI compliant.

Those 8,000-plus cards and 22,000-plus transactions matter, because according to Cisero's lawsuit (technically, a counterclaim to the card processor's lawsuit to collect the Visa and MasterCard fines on behalf of the bank), Visa's own rules don't allow recovery fines in cases where the number of card numbers involved in a breach is less than 10,000.

And how did Visa get from $1.3 million in "actual fraud" down to just an $80,000 fine?And how did Visa get from $1.3 million in "actual fraud" down to just an $80,000 fine? Here's what Cisero's lawsuit says:

On July 18, 2008—again without notice to Cisero's—Visa advised U.S. Bank that its ADCR review committee had reviewed the facts and had preliminarily determined that the alleged data breach qualified for ADCR processing. Visa alleged that the process was based on a review of 32,581 accounts claimed to have been stored on the Cisero's system. This number, which was not explained, differed considerably from Cadence's finding of only 8,100 account numbers and even from Cybertrust's count of 22,700 "instances" of Visa credit and debit card account numbers.

Using its own unexplained methodology, Visa then estimated the "actual fraud" caused by Cisero's non-compliance to be $1.26 million, a number which it then apparently adjusted based on a "baseline" of the ordinary amount of fraud across the Visa system. Visa arrived at an estimate of this "incremental fraud" caused by Cisero's non-compliance and added recovery for operating expenses for issuers, for a total of $521,600. Visa then "capped" U.S. Bank liability at $55,000, "assuming Cisero's Ristorante and Nightclub, and US Bank and any related agents, fully cooperate with the compromise investigation (e.g., providing information within the requested timeframes, demonstrating satisfactory progress toward remediation of PCI DSS violations)."

In October 2008, Visa performed its "final ADCR liability calculations." Visa calculated the "total event fraud" to be $1.33 million and Cisero's "total pre-cap liability" to be $511,513.41. Visa once again failed to explain how it arrived at these calculations or to provide any supporting documentary evidence.

A cynical observer might conclude that Visa didn't really decide to swallow a $1.26 million fraud loss out of the goodness of its heart—especially when Visa's accounting is way out of proportion to the documented chargebacks by MasterCard issuing banks, which came to a total of $13,850. (American Express didn't claim any chargebacks, even though there were more AmEx card numbers than MasterCards found on the Cisero's server.)

Cisero's undoubtedly non-cynical lawyer, on the other hand, called Visa's process "little more than a scheme to extract steep financial penalties from small merchants such as Cisero's for the benefit of Visa."

Did thieves actually steal card numbers from Cisero's? Is Visa's common point of purchase analysis trustworthy? Does Visa's accounting sound credible? If this case ever gets in front of a jury, some of those questions may be answered.

In the meantime, Cisero's and its Washington-lobbyist law firm, Constantine Cannon, are challenging everything from the contracts that acquiring banks require of their card-accepting customers to the ability of those banks to directly extract fines from merchant accounts, along with Visa's and MasterCard's ability to extract fines in a quasi-judicial procedure that doesn't allow a merchant accused of a security breach to present a defense.

Yes, those fines are officially fines for failing to comply with PCI requirements, and they're designed to be punitive. Cisero's lawyers are challenging that, too, along with the practice of acquiring banks to promise PCI support for small merchants and then failing to deliver—as well as the fact that to challenge a Visa fine, an acquiring bank would have to pay a non-refundable $5,000 fee to Visa.

In short, the lawsuit challenges what appear to be the most Kafkaesque elements of Visa's PCI enforcement process. How successful it will be is an open question. But one thing seems certain: With lobbyist backing, this is no ordinary lawsuit that's likely to be settled soon.

That could eventually mean real changes in the PCI-enforcement process—and it's almost certain to mean lots of ugly details from the underside of Visa and PCI.