The Legal Perils Of Cyber-Insurance For Retailers

Attorney Mark D. Rasch is the former head of the U.S. Justice Department's computer crime unit and today serves as Director of Cybersecurity and Privacy Consulting at CSC in Virginia.

There's an old saying in cyber-security: Identify the risks, mitigate the ones that are cost-effective to mitigate, and insure against the others. But a recent case out of an Ohio federal appeals court demonstrates that, while it may be easy to buy cyber-insurance, it may take the filing of a lawsuit to successfully file a claim once you have a loss. As a retailer, it may behoove you to take a good long look at all of your insurance policies, including the riders and exclusions. And in reading the policies, remember the words of Lewis Carroll's Humpty Dumpty: "When I use a word, it means just what I choose it to mean: neither more nor less."

In the aftermath of the hack of DSW's computer systems by uber-hacker Alberto Gonzalez, the Columbus, Ohio, shoe chain attempted to recover some of its losses by filing an insurance claim for the theft. The hack was relatively straightforward: For two weeks in 2005, Gonzalez accessed the unsecured wireless network at a single DSW store and, through that store, hacked into the retailer's core systems and stole payment-card numbers, account names and other information about 1.4 million customers at 108 different DSW stores.

In response to the breach, DSW incurred some major expenses, including the cost of investigating and mitigating the breach, notifying customers of the scope and extent of the breach, issuing credit reports, credit freeze and credit watch services for affected customers, and replacing credit, debit or other payment cards for hundreds of thousands of DSW customers, not to mention a series of lawsuits and FTC investigations that resulted in the institution of a compliance program and the payment of fines. The litigation alone cost DSW upwards of $4 million. DSW's calculated losses exceeded $6.8 million—all from an unsecured wireless router at one store. Well, mostly from an unsecured wireless router at one store.

But DSW was lucky. The chain had bought a "Blanket Crime Policy" from National Union Fire Insurance Company (NUFIC). That policy included a rider specifically covering claims arising out of "computer fraud." DSW promptly filed a notice of claim, and eventually filed a claim under the policy with NUFIC.

But NUFIC and DSW apparently read the policy completely differently. The insurance company claimed that the damages and losses were not covered. The policy covered "Loss which the Insured shall sustain resulting directly from the theft of any Insured property by Computer Fraud." NUFIC argued that there was no "theft" of any property because, look, the credit-card numbers, personal information, etc., are all still there! They haven't been "stolen" any more than your credit card is "stolen" when the clerk you hand it to looks at the number. Besides, NUFIC argued, the "losses" suffered by DSW were not "resulting directly" from the "theft" of the credit-card numbers but rather from the later misuse of those numbers. Only those losses that result directly from "theft" are covered.

The insurer also appeared to argue that the "property" stolen—consumers' credit-card numbers—was not the "property" of DSW but instead belonged to the customers or to Visa/MasterCard. Because most of the costs were paid to third parties (the banks that reissued consumers' credit cards) or on behalf of third parties (credit fraud watch services to consumers), these losses were "indirect" and, therefore, not covered by the policy. NUFIC argued that the phrase "resulting directly from" computer fraud refers only to the insured's own loss and not to the insured's vicarious liability to third parties (like banks and consumers), and it relied on a long line of Ohio and other cases that discussed fidelity bonds (covering employee dishonesty).

The Ohio Court rejected this approach.The Ohio Court rejected this approach and applied something not generally used in interpreting insurance contracts—common sense. It said, "directly related to" computer crime meant "caused by," as opposed to "indirectly caused by" or "consequential damages" or too far removed from the computer crime as to be reasonable. So the costs of replacing a credit card "stolen" electronically is directly related to the theft, but the cost of installing new routers may or may not be—depending on how directly it is related to the actual theft.

As to the argument that the records were not "stolen" because they were still there, the court noted that the policy provided coverage for loss that the insured sustained "resulting directly from" the "theft of any Insured property by Computer Fraud," which includes the "wrongful conversion of assets under the direct or indirect control of a Computer System by means of fraudulent accessing of such Computer System." The consumers' payment-card numbers and other information were certainly under the control of the computer system and were "converted" by Gonzalez and his gang through their fraudulently access to the computer system.

But wait, there's more.

NUFIC had another reason for not paying the claim. You see, the policy excluded claims resulting from "loss of proprietary information, Trade Secrets, Confidential Processing Methods, or other confidential information of any kind." Because the payment-card numbers were "proprietary" or "confidential" information, and that information was "lost," no claim could proceed. If there were no coverage for the loss of the information itself, the insurance company argued, there would also be no coverage for the economic damages resulting from the loss of the information.

The insurance company relied on the fact that DSW treated customers' data as confidential, that it was required to protect such data under Visa, MasterCard and PCI DSS rules, and that the chain submitted statements to the FTC asserting the statements (which included payment-card information) contained "confidential information" as proof the claim arose from the loss of confidential information and was, therefore, excluded. Catch-22.

Not so fast. The court noted that to interpret "other confidential information of any kind" as the defendant urged—to mean any information belonging to anyone who is expected to be protected from unauthorized disclosure—would swallow not only the other terms in this exclusion but also the coverage for computer fraud. Although the customers' credit-card numbers were certainly sensitive and protected from misuse, they were not the types of trade secrets that this provision was likely intended to exclude from protection.

All of this goes to show that, even if you buy specific cyber-insurance, there is a huge difference between what you think it covers and what your insurer thinks is covered, particularly after you file a claim. Thus, all retailers should look at all of their insurance policies for language that might be used to deny claims and then figure out what is and what is not covered in advance. And insurance companies need to do a much better job explaining this stuff in clear language if they expect to be able to sell these policies in the future. Let's not forget how Alice answered Humpty Dumpty's claim that words mean what he says they do. "The question is," Alice said, "whether you can make words mean so many different things." That, my friend, is a job for a lawyer.

If you disagree with me, I'll see you in court, buddy. If you agree with me, however, I would love to hear from you.