Is Sony Making The Business Case For PCI?

A 403 Labs QSA, PCI Columnist Walt Conway has worked in payments and technology for more than 30 years, 10 of them with Visa.

The PCI Council has stated that compliance is cheaper than noncompliance. The idea is that security is a good investment and that, although meeting all the PCI requirements may be expensive in terms of time, technology, management attention and money, the alternative costs more. The problem with such a statement is that, although it sounds good, we do not have much evidence to support it.

That situation may be changing with the recently announced data breaches at Sony. Its experience could provide the business case for why every retailer should want to be PCI compliant. Some large retailers have been breached in the past, but their customers eventually came back. Many of these customers may even have felt they were more secure after the remediation efforts undertaken by the merchant. We've even seen a major processor suffer a data breach, but it's still going strong.

In short, those of us preaching PCI compliance did not have much evidence to support our contention that PCI noncompliance really is more expensive. Sure, there were fines and some brand damage. But the effects seemed transitory and the businesses moved on. All that has now changed, thanks to Sony's unfortunate experience.

What we know from public announcements so far is that person or persons unknown hacked into Sony's PlayStation Network servers in mid April. They took personal information on account holders, including names, addresses, E-mail addresses, user IDs and passwords for about 77 million customers. Subsequently, we learned the breach extended (maybe the same breach, maybe the same hacker—we don't know) to Sony Online Entertainment (SOE) and an additional 24.6 million customers.

From a PCI point of view, the company reports the hacker(s) also may have accessed the credit-card numbers for at least a good portion of those 77 million Sony PlayStation users, plus another 12,700 credit and debit cardholders outside the U.S. who were SOE users. Sony claims the credit-card information was encrypted and did not include the security codes, although the details are still vague concerning those 12,700 non-U.S. accounts.

Sony is paying a high price for the data breach. Aside from all the usual bowing and apologizing executives, the company's stock price recently dropped to its lowest level in two years. Sony also can expect to take a direct financial hit, because it had to shut down the PlayStation Network. My understanding is that the division contributed a disproportionately large share of total corporate profits, so its closure and customer alienation is more than just an embarrassment (and likely accounts for the falling stock price).

The wild card is whether the hacker(s) will be able to decrypt those 77 million (or however many there are) credit-card numbers. I can only hope that Sony used strong encryption (as required by PCI DSS—more on this later) and that the hacker(s) did not gain access to the encryption key material. Can Sony take that chance?The company does not have a lot of good choices. It could decide to trust its encryption, assume the keys were not compromised and do nothing. The risk is that the hacker(s) or someone even more sophisticated breaks it and gains access to about 77 million cleartext PANs. The potential fraud losses for which Sony could be liable—pick your favorite number: $25, $50, $100 per account—would be devastating to the company.

Alternatively, Sony could try to get ahead of the curve and recommend that users contact their banks and cancel their cards. In this case, the issuers are likely to have no sense of humor, and they will want to be compensated. Using a very rough estimate of $10 to $20 to re-issue a card, this cost would approach or even exceed $1 billion. Paying for credit monitoring instead may be a third option. However, I doubt it will be much cheaper, given the large number of users who would need to be covered.

What we see so far is a dropping stock price, measurably lower corporate profits and brand damage in the eyes of customers that could crimp revenue and profits for years and that may, ultimately, be unrecoverable. In other words, we are gathering hard data for the costs of a very large and very visible breach and potential data compromise.

Was Sony PCI compliant at the time of the breach? We do not know, and I expect we will not know for a while. The answer will have to wait until we learn the outcome of the forensic investigation that is sure to happen (trust me, losing 77 million accounts will get the card brands' attention, and they and Sony's acquirer will want some answers).

Everyone, including the individuals affected, needs to know that Sony properly encrypted the payment card database (PCI Requirement 3.4), that it had compliant key management processes (Requirements 3.5 and 3.6), that it did not store the security codes as reported (Requirement 3.2) and that it minimized its cardholder data storage (Requirement 3.1). I'm particularly interested in that last part, given the magnitude of the breach, and I don't know how many others are paying attention to it.

As we follow the slow-motion train wreck that is the Sony PlayStation data breach, it is easy to miss the bigger data security and PCI lesson that applies to every retailer. That lesson is: If you think PCI compliance is expensive, noncompliance can cost even more. Another lesson: Every retailer should re-examine its risk assessment (PCI Requirement 12.1.2) and ensure it reflects the company's changing business realities and vulnerabilities. For what it's worth, this is the same advice I offered after the RSA breach.

I have always agreed with the PCI Council that PCI is a worthwhile investment and that the cost of noncompliance is higher than the cost of being compliant. With the unfortunate experience of Sony, we may have more hard evidence to support that argument.

What do you think? I'd like to hear your thoughts. Either leave a comment or E-mail me at [email protected].