My company permits remote access with authentication that includes:

  1. username
  2. password/phrase (20+ characters)
  3. User selected PIN + RSA SecurID token code

Our passwords are required to be changed every 6 weeks, and of course the SecurID token changes every 60 seconds.

Recently the policy has been modified slightly requiring the PIN to be changed every 12 months (there was previously no requirement to change the PIN). I can't personally see how this increases security, but I'm not an expert.

What security benefit does a PIN change policy like this give, and more specifically (if possible), what threat / attack is mitigated by this?

  • Is the new policy a change from "your PIN is permanent", or a change from "you need a new PIN every six weeks"?
    – Mark
    Commented Jun 6, 2017 at 23:27
  • The change for the PIN is from "permanent" to "change every 12 months".
    – Sam
    Commented Jun 6, 2017 at 23:35
  • 3
    It sounds like the new policy mitigates the threat of management being told they're not doing enough to mitigate threats.
    – Mark
    Commented Jun 7, 2017 at 0:16
  • 3
    Such policies seldom make sense, don't waste brain cycles on it. It's designed to get a checkmark on some assessment form. It might fulfill the function of detecting unused tokens (If pin expires)
    – eckes
    Commented Jun 7, 2017 at 0:17

1 Answer 1


Nothing really, if your environment already employs a 20+ character password then this PIN is kind of "doubling up" on this. The RSA tokens are usually deployed as a one stop solution. You log in with your user name, you know the password and you have the token (which gives a code). The trinity of who you are, something you know, something you have.

You have the extra layer of a strong password which is stronger than then pin used for the token.

I mean, correct me if I'm wrong? but i think the only thing that this protects against in these conditions are lost keys or old keys out in the wild that are unknown for some reason. Also an end of life process that is automatic, as the keys that are more than a year old will simply not be useful to anyone down the road.

Some of the more paranoid sysadmins might think if an attacker gets a key and knows who its from it might forever be a risk, since all they have to do is find out the rest of the details, having no end in sight to that system gives the attacker unlimited time, in theory, to do recon. This rule would provide a one year constraint on finding out the rest of the details required. A sort of automatic garbage collection if you will. But that's the only thing i can think of here.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .