Would they be able to get around this by encrypting it two times? On of the times with the official recommend standard an another time with bcrypt?

https://www.pcisecuritystandards.org/pdfs/pci_dss_glossary_v... seems to list Blowfish as an approved standard as far as the PCI standards group is concerned.

Thanks for that link. I think that means blowfish is approved as a standard encryption algorithm, but that doesn't mean it is approved as a standard way to store passwords.

From http://security.stackexchange.com/questions/4781/do-any-secu..., I gather this: bcrypt is neither a NIST or FIPs standard, but PBKDF2 is.

> For any business required to comply with U.S. NIST or FIPS standards, bcrypt is not a valid option. Check every nation's laws and regulations separately if you do business there, of course.

However, from a comment on this other answer: http://security.stackexchange.com/questions/11552/would-it-m..., someone mentions this:

> As far as I know, PCI doesn't specify actual technology, especially Encryption specifications. The legal cases I've seen support what's technically feasible, not technically ideal or perfect. I'm sure either one [bcrypt or PBKDF2] meets PCI requirements

An opinion shared by many people that deal with PCI, is that you follow PCI not to secure your systems, but to limit your exposure to PCI fines and remedies. It seems like it is a rock-solid defense for PCI compliance to choose the algorithm that is certified by NIST.

But it is also likewise plausible to say that you're using a newer technology that's understood by experts in the industry to be even better than those certified some time ago.

I agree with you, but there is zero risk w/r/t PCI:DSS if you use the FIPS standard.

The difference is that one way your lawyers won't even need to make an argument and the other way they might have to make an argument.

Would it matter? I mean all they have to do is encrypt it using the old standard and then encrypt that using bcrypt.

Lawyers "do you use blabla standard to encrypt the data?" You> "yes" - and it is secure.

Could you double hash passwords? First with bcrypt for actual security, and then a second time with SHA256 as a bureaucratic measure?

I have a link above about about using PBDKF2 and bcrypt together http://security.stackexchange.com/questions/11552/would-it-m...

The user Thomas Pornin knows his crypt which is why he has 86k reputation.

That's quite an assumption.