There are a lot of rumors about rules that companies must abide by security-wise to store banking information - say, checking account and routing numbers. It's very difficult to dig up anything factual on the matter, though.
It seems to me that it must work one of three ways:
1.) The government provides a set of security standards which must be met to legally store banking information.
2.) Banks provide security standards which must be met to retain their business.
3.) There are no specific standards, but you are liable to be sued for any breech in security.
I know how important it is to have near impenetrable security when storing important information. I'm looking specifically to demystify the legal end of things. Have I hit the target with one of these three theories?
PCI-DSS is the most commonly used standard, is aimed at retailers and payment processing systems. And while it is credit card based, much of what is in it covers other stuff that you should be thinking about if you're storing banking information.
One book to look at is Cryptography in the Database. There is a section about laws that cover data security such as GLBA (which says nothing that a developer finds useful) and SOX (which, for software development, is more about background checks and version/configuration control). http://www.amazon.com/Cryptography-Database-Defense-Symantec...
Another book that may help with keeping the data away from hackers (and rogue employees) is Translucent Databases. I have the 1st edition, and the 2nd just came out last month: http://www.amazon.com/Translucent-Databases-2nd-authenticati...
In support of #1, check out NIST's 800 series of standards. When we were looking to bid on a government computing contract, they included a long list of them by reference, effectively turning a 3k page RFP into about 6k pages: http://csrc.nist.gov/publications/PubsSPs.html