Friday, March 12, 2010

RSA 2010 EXCLUSIVE PCI Security Standards Council Interview

At RSA 2010, I was given a unique opportunity to interview Bob Russo (GM at PCI SSC) and Troy Leach (CTO at PCI SSC). I have prepared a deck of very tough questions and then had an hour-long discussion with Bob and Troy around those questions. What follows is the interview reconstruction from my notes with minimum edits and clarifications by the Council folks.

Anton Introduction:  I think PCI DSS is the most valuable thing to hit security industry since its inception – both as a driving force for security improvements and as a source for security guidance. However, there are skeptics among merchants (too much security) and some security professionals (too little security). Some of my questions below focus on dispelling the misconceptions such skeptics might hold.
Anton Question 1: What, in your opinion, is the main value of PCI DSS – to the community at large? Merchants? Banks? Brands?
Bob and Troy @ PCI Council answer:
You have answered this question yourself above: it is security. Motivation for payment security improvements is the value of PCI. For some companies it is also a springboard for additional security improvements needed for their businesses. This benefits everybody!
PCI value can also be rephrased as demonstrating trust across organizational boundaries and. As we know, payment security has many sides and PCI compliance is one way of demonstrating trust across organizational boundaries.

Anton Question 2: Way too many companies seem to focus on compliance and not on security. What is the best way to prevent “teaching to the test” for PCI DSS compliance?
Bob and Troy @ PCI Council answer:
Too many companies focus on studying for the test. We believe the PCI Standards provide a solid foundation for a security strategy to look after payment and other types of data, but security does not start and end with compliance with standards.
Education is very important and that is why the PCI Council will focus even more on educating the merchants and changing their mindset from one of compliance to security. Their old way of doing business – retaining card data, for example- was viable one day, but not today.One of the steps we see is increased outsourcing of payment processing to trusted providers.

Anton Question 3: Some people say that “the brands must just change the system” since Level4 merchants [=typically smaller merchants] can never be educated and this never can be secured. What do you say to this?
Bob and Troy @ PCI Council answer:
It’ll happen eventually, but it is obviously not so easy. We’re talking 5 to 10 years here. The payment system is diverse and incredibly complex. Any drastic changes will probably be more costly and disrupt merchants’ business even more than PCI DSS ever could, so they have to happen gradually. The PCI Council’s mandate is to get as much done to improve payment security as possible - within the existing system. Security has to become part of every business that deals with card data.

Anton Question 4: There are many debates about PCI DSS in security industry, among merchants, etc. How can the impact of PCI DSS payment security be measured? Who might have the data to do it?
Bob and Troy @ PCI Council answer:
Security breach statistics demonstrating a root cause that can be mapped to PCI DSS requirements is one such possible way to prove the value of PCI. For example, if the company did not take any measures to protect against SQL injection and got breached through that, they need to pay more attention to Requirement 6.6.
On the other hand, trying to analyze what the non-breached companies are doing right with PCI is harder since you don’t hear about the myriad of success stories of companies that are defending against breaches through following DSS or have minimized card data compromise in breach situations through strong logging and monitoring, mandated by PCI.
PCI DSS prescribes logging and monitoring, which help detect data loss. Unfortunately some recent incidents had breach evidence present in logs, but since logs were not reviewed until breach became public (contrary to PCI DSS requirements) this was not utilized for detecting the breach.
More education efforts are needed to explain to merchants that PCI is not only about breach prevention, but also about detection of intrusions and security monitoring. Thus, judging its value only on breach prevention is shortsighted.
Enhanced information sharing will drive more improvements here.

Anton Question 5: What is your opinion of mandating the discovery of stored card data and especially track and other prohibited data? This technology was not high on the list in PWC report.
Bob and Troy @ PCI Council answer:
Many QSAs already use data discovery tools today. Since PCI scope covers systems where card data is present, payment card data discovery should be part of scope validation. “Forgotten” credit card data dumps were indeed present in some recent breaches stories.
Methods of such discovery can vary- using an automated tool is one of the options, but such tools are still not mature.

Anton Question 6: Do you think that there should be tiered security requirements for small and large organizations (that go beyond today’s SAQ validation levels)? For example, daily log review seems onerous to many merchants.
Bob and Troy @ PCI Council answer:
You cannot dumb security down below a certain level. More education efforts will be needed to explain to merchants how to satisfy requirements and become compliant [and stay compliant].
However, the Council is planning to build more tools in order to help merchants understand what exactly they need to do to become compliant. A wizard interface or some other method to simplify the SAQ process can be used here to highlight which controls the merchant needs to implement.

Anton Question 7: The “None were compliant when breached” rings true to me. Why do you think so many people object to this?
Bob and Troy @ PCI Council answer:
People simply need to know the facts and find out what happened in those breach stories. For example, some breached companies had massive stores of prohibited data, such as authorization data. Or they were not adequately protected at the application or database level against things like SQL injections. There is a difference between “breached due to negligence” and “breached due to bad luck.” Being diligent but still ultimately failing to protect the information is possible (so safe harbor does exist for such companies); it just isn’t what happened in those incidents.

You just need to get the facts. If a company gained compliant status by misrepresenting the facts to a QSA, PCI standards are not at fault when the breach happened.


Anton Question 8: What is the best way to balance PCI DSS lifecycle with both merchants complaints about “moving target” and with rapidly changing threats?
Bob and Troy @ PCI Council answer:
So far, the current two year lifecycle has provided a good balance between structured development and staying abreast of rapidly changing threats. Feel free - and have your readers - to suggest changes to that lifecycle, if you think it needs to be changed! We are considering how it might evolve.

Anton Question 9: What do you think of using PCI DSS controls for non-payment-card data?
Bob and Troy @ PCI Council answer:
It is a good thing, if you keep in mind that PCI DSS controls are the foundation or the minimum baseline for an effective security strategy. Organizations will likely need to build more security on top of the PCI foundation to protect other sensitive data. Layering technology solutions and combining with the necessary people and processes continues to be the most effective means in protecting cardholder data.
PCI has certainly raised awareness for all data protection, not just payment card data.
Anton Summary
Overall, the main themes I picked in the conversation were:
  • “PCI compliance” is a means to an end. And the end is “security!”
  • Education is one of the ways to change the thinking of merchants and to improve security.

Thanks to Bob and Troy for the insightful discussion!

Dr Anton Chuvakin