JCPenney's Breach: Differences From Feds, Gonzalez, JCPenney Itself

In November 2007, Albert Gonzalez's crew was in the midst of hitting their laundry list of major retailers when they used their SQL attack on JCPenney. But just how deep they penetrated the $18 billion clothing chain is unclear, with the Justice Department, JCPenney and intercepted messages from Gonzalez IM conversations all painting very different pictures.

Gonzalez has pleaded guilty to placing a file called sqlz.txt—contained information from JCPenney's network—on a Ukrainian server. Assistant U.S. Attorney Stephen Heymann said the government wasn't sure what Gonzalez's team got from JCPenney: "Although the protective system used by JCPenney had unquestioningly failed, the Secret Service had no evidence as to whether payment card numbers had been stolen."

As for Gonzalez's comments at the time, a series of ICQ messages in November and December 2007 gave some hints of the perceptions—at that time—of the cyberthieves. A Nov. 1 exchange said they were searching for weak passwords at the chain and that access wasn't difficult: "They have most of (their) ports open. Wasn't too hard."

On Dec. 16, Gonzalez then asked his associate: "I'm curious how Hacker 2 moved around on jcp so quickly w/o making noise" and he was told "SQL Servers is his key to everything. Heh." On Christmas Eve, Gonzalez declared "I got access to the JCP POS network."

The government now says that Secret Service agents alerted JCPenney in May 2008 "that it had been attacked in or about October 2007. Prior to receiving that alert, (JCPenney) was wholly unaware that it had been previously victimized by computer hackers."

(Lawyers and a federal judge go a few rounds over trying to keep JCPenney's name out of the case. The Judge wins.)

Side note from various federal documents that were unsealed this week: If anyone had been trying to visit 7-Eleven's site on Nov. 9, 2007, they would have discovered the site down. But it wasn't a software glitch or excessive traffic. It was the convenience store chain's attempt to block Gonzalez's attack. "7-Eleven disabled its public-facing Internet site to disrupt the unauthorized access," the government filing disclosed.

Back to JCPenney. The chain's version of its attack is that credit cards were indeed taken, but only two of them, and both were the chain's own branded cards. Most importantly, though, when the cyberthieves thought they had broken into JCPenney's centralized POS database, they had actually penetrated something much less valuable: "The hackers did not access JCPenney’s POS network, as was assumed by the hackers in their chatter, but rather a POS server in a single store, where they retrieved two private label credit cards," said Darcie M. Brossart, VP for corporate communications at JCPenney.

Brossart said the chain considers none of the data to have ever been at risk, given that the thieves never got close to the core payment card systems. "They had been in the system but they weren't getting anywhere. They were nowhere near the POS system, they were nowhere near penetrating actual customer databases," she said. "There was never a risk of customer information being revealed."

"Based upon the detailed forensic investigation that was completed on JCPenney’s systems, we know exactly the methods employed to gain access into our system, the tools used and the probes conducted against machines in our network," Brossart said. "Our assumption is that once the hackers accessed the SQL server, the complexity of our network and its set-up, made it extremely difficult for them to compromise data."Brossart makes some very valid points about the distance that the thieves maintained from the most valuable data. But it's hard to take the leap that "there was never a risk of customer information being revealed," given that the team did break into the network and was able to access two credit cards completely enough that the chain chose to replace those customers' cards.

Federal filings show that JCPenney retained IBM and Mandiant to help with its internal investigation of the intrusion.

The situation is a little less clear with the most newly identified victim of Gonzalez: $561 million chain Wet Seal. Ed Thomas, Wet Seal's CEO, issued a statement on Monday (March 29) confirming that it had been breached but declaring that the thieves got nothing. "We are pleased that time has proven, as we believed from the outset, that none of our customer information was taken," Thomas said.

The U.S. Probation Department's pre-sentence report sees it somewhat differently: "Forensic analysis reflects that Gonzalez was working with data stolen from [Wet Seal] less than three weeks prior to his arrest." And Gonzalez's plea agreement said "on or about April 22, 2008, Gonzalez modified a file on the Ukrainian Server that contained log data stolen from [Wet Seal's] computer network."

Government lawyers have said they can't prove that any Track 2 card data was taken from Wet Seal, so the chain could have said that. Not that none was taken, but that the government can't prove that any was taken. Wet Seal could have even said that no credit or debit cards were taken, although the government didn't go that far. But to say "no customer information was taken" seems at odds with the government filings. Either that or Wet Seal is taking a very narrow view of customer information. What would the cyberthieves have taken that wasn't customer data? Even log files of purchase activity is customer data.