Educause Security Discussion mailing list archives
Re: stopping students sharing their login credentials
From: Neil Sindicich <neilsin () KSU EDU>
Date: Fri, 23 Jan 2009 10:48:12 -0600
Suffice it to say, that both things may be necessary. One policy stating: "Don't share your password" And another stating: "You are responsible for what happens on your account." However in the case of speeding in cars (and the analogy may eventually be commutable) many states have outlawed the use of speeding cameras, or speed based toll systems because it was considered unfair/unconstitutional(in that state) to hold a car owner responsible for the person who borrowed or even stole their car. That said, the key to any policy is the awareness of the people who are required to follow the policy. You get the tree in the forest analogy, I think it goes: "If a policy is posted on a tree in the forest and no one is around to see it, does the policy need to be followed?" All of us would like to think that the policy would need to be followed (/Ignorantia legis neminem excusat/) but the truth of the matter is, this is a really hard ideal to enforce. Making a policy for "Don't share your password" means that when you do find someone who has shared their password (probably because something was done with it that should not have been) then you have recourse to take disciplinary action of some sort. You could still take disciplinary action based on the "You are responsible for what happens on your account" but will you if the account information was obtained because of bad security practices somewhere else, or a breach of a server, or some other circumstance that you haven't discovered yet? Probably not... In either case the first question you will probably ask the individual in question is: "Did you give out your user id and password to anyone?" and then we have to figure out whether or not they are lying... Cheers, Neil Sindicich Cyber-Security Analyst Kansas State University email: neilsin () ksu edu phone: (785) 532-2598 randy marchany wrote:
Don't get me wrong. There needs to be some sort of "control" in a policy or standard. The control has to be enforceable,however. If not, people will ignore the policy. The "personal responsibility" item provides that control. Your speed limit example is good but given the information available in computer logs, the speeding violation would be "Randy's car was speeding but we don't know who the driver was" so we can't give a ticket out because we can't identify the individual who was actually driving. Remember, computer login logs (your only evidence of account login) only show the userid and not the person. At least, with the responsibility clause, it'd be "Randy's car was speeding and Randy is responsible for its use and therefore accountable". Car insurance companies follow this strategy to some degree. The computer logs would show Randy's userid was used and the policy states that Randy is responsible for its use. You have an actual person to "interview". -r. On Fri, Jan 23, 2009 at 10:52 AM, James M. Dutcher - Assoc. VP IS/IT & CIO <james.dutcher () sunyorange edu> wrote:Randy, et al, I would like to offer a counter point. I believe that there has to be a policy in place. Otherwise, anyone can contest that "they did not know" or " you did not say that I could not do it". Having a policy protects the organization. Yes you are correct that it is difficult if not impossible to police/enforce, especially in real time. However, when there are digressions encountered/discovered, then the appropriate actions take place and the diggressors are then the examples (and precedents) as to what happens when policies are broken. Take for example highway "speed limits". There is not enough police/surveillance in place to ensure that everyone complies with it. But there is some in place to catch folks so as to (hopefully) keep the rest of the drivers in compliance. Regards, Jim James M. Dutcher - PMP, CISSP, SCP/Security+, CISA SUNY Orange - Associate Vice President of Info. Tech. Services & CIO 845-341-4651 (office) 845-742-8954 (college cell) 607-760-7455 (personal cell) james.dutcher () sunyorange edu jim () dutcher net Yahoo IM: jmdutche Google Talk: jmdutche On Fri, Jan 23, 2009 at 10:32 AM, randy marchany <marchany () vt edu> wrote:One should never put in a policy/standard any item that can not be enforced. While the spirit of the statement "you must not share your userid, login credentials with anyone" is certainly clear, the reality is that this cannot be enforced without additional monitoring such as 2 factor authentication, video feeds or witnesses. The most common "abuse" of this would be a simple login and computer logs cannot show that a userid was shared or WHO the person was that actually logged into the system. Biometrics isn't foolproof since I could login with my biometrics and let you use my userid. Card swipe access isn't foolproof since people form trains to enter a facility. Shoot, I remember visiting a campus, going to a pizza place right across the street and seeing the building access code written on a piece of paper on the bulletin board next to the cash register. Apparently, that pizza place delivered a lot of pizzas to labs in the building :-). So the "must not share" clause is basically unenforceable and weakens your policy/standard. Another way to express the intent of the clause is needed. Why do people share these things? Could be something as simple as the site doesn't have a mechanism for guest access. An example of this is guest wireless access on campus. You have a guest speaker who needs wireless access, your campus has no mechanism to provide guest access so you, the sponsor, lets the speaker use your credentials to get access. Email access, door access are other examples. An email system doesn't allow sharing of email folders, a dept. head is on travel and the assistant needs access to those emails. The only alternative is to share the email password. So, how do we fix this? The best solution I found was to state "you are responsible for whatever activities originate from your userid, computer, id card..." (feel free to include whatever authentication/authorization mechanism you have). This is easily enforceable. Computer and access control logs note the "userid/token" that was used to gain entry. SInce you can identify the owner, that person is responsible for its use. I do believe having the "responsible for its use" strategy is more effective. In Russell's case, the access logs contain the name of the card owner. You contact the card owner and ask them the necessary questions :-). Just my .02. Randy Marchany VA Tech IT Security Office and Lab On Thu, Jan 22, 2009 at 9:25 PM, Russell Fulton <r.fulton () auckland ac nz> wrote:Background: Earlier this week we had an incident where the building security officer noticed a group of unfamiliar people using machines in one of our labs. She asked them for their ID cards and none could (would?) produce one. On questioning they said they were students from a neighbouring institution and that they were using "borrowed" credential. We have cctv footage and swipe card logs from the door (which may show they tail gated someone in). We are now tracking down which machines were being used so we can disable the accounts. To the point. We (the security techies) have been asked what measures we can deploy to prevent this sort of thing happening in future. We already do lots of education, posters, page on the back of the student handbook. Students have no excuse for not knowing that they should not share passwords. On the social/education side we could make an example of anyone we finger for this (assuming we can make charges stick) in the hope that this will persuade other students not to share their passwords. Technical solutions seem to revolve around some form of two factor authentication. I.e. something the student has but which they will be reluctant to part with for any length of time. Like their ID card. Our ID cards have bar codes and classic mag stripe. Some labs (like this one) also have proximity card locks. Generally only post grad students or students in special coursed (like medicine) have proximity cards. Anyway I would very much like to know what other are doing in this space. Cheers, Russell
Current thread:
- Re: stopping students sharing their login credentials, (continued)
- Re: stopping students sharing their login credentials Rizzo, Jim (Jan 22)
- Re: stopping students sharing their login credentials Ray Strubinger (Jan 23)
- Re: stopping students sharing their login credentials Rappaport,Jason (Jan 23)
- Re: stopping students sharing their login credentials Mike Wiseman (Jan 23)
- Re: stopping students sharing their login credentials randy marchany (Jan 23)
- Re: stopping students sharing their login credentials James M. Dutcher - Assoc. VP IS/IT & CIO (Jan 23)
- Re: stopping students sharing their login credentials Christopher Jones (Jan 23)
- Re: stopping students sharing their login credentials randy marchany (Jan 23)
- Re: stopping students sharing their login credentials Mike Wiseman (Jan 23)
- Re: stopping students sharing their login credentials Charlie Reitsma (Jan 23)
- Re: stopping students sharing their login credentials Neil Sindicich (Jan 23)
- Re: stopping students sharing their login credentials Barros, Jacob (Jan 23)
- Re: stopping students sharing their login credentials Basgen, Brian (Jan 23)
- Re: stopping students sharing their login credentials Brad Judy (Jan 23)
- Re: stopping students sharing their login credentials Charlie Reitsma (Jan 23)
- Re: stopping students sharing their login credentials Gary Flynn (Jan 23)
- Re: stopping students sharing their login credentials Valdis Kletnieks (Jan 23)
- Re: stopping students sharing their login credentials Neil Sindicich (Jan 25)
- Re: stopping students sharing their login credentials James M. Dutcher - Assoc of IS/IT & CIO (Jan 25)