PKI versus raw public keys

At the X9F4 meeting last week, we started work on a new document that will define a set of criteria for the secure use of PKI by financial institutions. There was unanimous agreement that the document needed two major parts: one that covers PKI that uses digital certificates and one that covers PKI that doesn't. Identity-based encryption is an example of using public-key technology in a way that doesn't require certificates. Using raw public keys also does. I've seen lots of use of IBE, of course, but I've also seen a few uses of raw public keys. Others in the X9F4 working group had also seen examples of that, and it seemed to me that the uses that we'd seen of it fell into two general categories.

In one case, if you're protecting things that are of relatively low value, you might decide that using certificate-based PKI just isn't worth the cost and headaches that it can cause, but you still want the benefits that using public-key technology can give you. I've seen cases of where exactly that happens.

In the other case, there are also cases where people just don't trust a certificate-based system to protect things which are of extremely high value. I've never seen examples of that myself, but other working group members had, and I have to admit that I was a bit surprised by these particular use cases.