[Cryptography] Paper check security

[Arnold Reinhold]

> On Oct 1, 2015, at 8:52 PM, John Levine <johnl at iecc.com> wrote:
> 
> [ not very much about crypto ]
> 
>> I think most consumers in the U.S. get their checks from check printing firms and I believe they generally include standard security features ...
> 
> Sure, but a check you print on your inkjet is just as valid.  Since
> everything is scanned optically, banks don't even care about magnetic
> ink for the numbers at the bottom any more.

Right, but I still have the option of using professionally printed checks with all the cool paper security features. They are available in sheets designed for laser printing too. And I think most people and business do use pre-printed checks. 

[crypto-relavance on]

But if the software that people use to print their own checks added a scan-surviving cryptographic signature that included the core information (bank routing, account number, check number, date, payee, amount and whatever I’ve left out), then I argue a plain paper inkjet check would be more secure than one printed on a fancy form but without the cryptographic signature.  Note that unlike a lot of cryptographic proposals, this would be easy to implement and would not require any time-consuming standards making process to get started. One software vendor, e.g. Quicken, could pick a format and start using it. Banks would ignore of course, at least until sufficient customer demand emerged. A simple app could be used to verify the sig.  If a different standard emerges later, it can be used on new checks without need for backward compatibility. 

Arnold Reinhold