Sony says data is protected, attackers say it's for sale

Claims from Sony and the PSN attackers seem to conflict. Sony says data was encrypted, but attackers are offering it for sale

Comments

Sony executives Kaz Hirai, president of Sony Computer Entertainment (left), Shinji Hasejima, CIO (right), at a [[artnid:384903|Tokyo news conference]] on May 1, 2011

Sony tried to calm customer fears by stating that the credit card data was encrypted, but attackers claim to already be selling that credit card data online. Either one of these parties is stretching the truth, or encrypting data doesn't offer the level of protection we think it does.

In a blog Q&A about the Playstation Network breach, Sony states, "The entire credit card table was encrypted and we have no evidence that credit card data was taken." Sony goes on to claim that it never collects the three-digit CVV number from the back of the card, but later amended that claim to state that it does collect that information, but it does not store it.

Meanwhile, the attackers claim to have the credit card data -- including the CVV number -- for sale on online black market forums. One of these things is not like the other. It seems difficult -- if not impossible -- to justify how both parties can be telling the truth.

Unfortunately, it is actually feasible that the data could have been encrypted as Sony claims, yet compromised as the attackers claim. It all depends on how the data was encrypted, and how the attackers breached the Sony network.

AppRiver security analys Troy Gill clarified that nothing is really known at this point, but added that if the data was encrypted as Sony claims, it is still possible that the attackers could have cracked that encryption by now. "It would depend first on what hash function was used to encrypt the data, obviously if a weaker encryption was used then the easier it would be to break. The amount of resources the hackers were using to break the encryption could also be a factor in the amount of time it would take."

Anton Chuvakin, security expert and co-author of PCI Compliance, notes that database table encryption is often poorly implemented. Organizations often use hardcoded encryption keys that an attacker might easily find once they have access to the network in the first place.

Tim 'TK' Keanini, CTO of nCircle, explains, "If the encryption keys were stored unencrypted somewhere convenient to the system it could also have been convenient to the hacker." Keanini also stressed that all we really have to go on right now are assumptions, but pointed out that even an unsophisticated attacker can figure out how to bypass or circumvent weak security practices.

Anup Ghosh, founder and chief scientist of Invincea explained that the fact is that even encrypted data is accessed and used internally, and the decryption typically happens seamlessly in the background for authorized systems or users. So, while encrypted data may be gibberish if extracted directly from the database, an attacker only needs to find the right system on the network to be able to pull the unencrypted information from the database.

All of these points address the feasibility or likelihood that Sony could be telling the truth about encrypting the data, and yet that data could be decrypted and available on the black market. However, we still have the issue of Sony claiming not to have stored the CVV data from the credit cards at all, and yet attackers claim to have that crucial piece of data as well.

If that turns out to be true, Sony could be in big trouble with the PCI-DSS powers that be. Chuvakin says that the PCI-DSS guidelines are very clear about the CVV data -- the three-digit code from the back of the credit card can never be stored in any form, whether encrypted or unencrypted. Attackers may have intercepted CVV data in transit rather than acquiring it from the database, but if that information was stored in the database Sony has some explaining to do.

Still, if Sony is telling the truth about encrypting the data, it seems it wasn't encrypted very well. It is also possible that attackers are bluffing, or flat-out lying to try to find a sucker willing to pay them for data they don't really have. Only time will tell.