AnsweredAssumed Answered

Use BrowserCheck for during a website's login process?  Mitigate the End Node Problem

Question asked by sweerek on Apr 19, 2012
Latest reply on Apr 25, 2012 by Eric Perraudeau

Hello,


Here's an idea using BrowserCheck to mitigate the End Node Problem (http://en.wikipedia.org/wiki/End_Node_Problem).

 

Let's say your website contains senstive info and your employees login into it to access/view/use that data.  Most of your site's visitors are good guys, but their devices vary in security.  Most end node scanning / validation assumes the user/device is malicious and tries hard to make sure it happens, happens right, and must pass -- some even foolishly try to force malware scanning.  But, as noted, most visitors are really good guys but often unknowningly entering with risky systems.  Since they're good guys (and employees w/ vested interest in thier group) they'd help keep the data secure... if they knew what to do and the 'cost' of doing so was minimal. 

 

The group's data owner trys hard to protect his data, but he must still give remote access.  He has hours of mandatory annual employee training, free anti-virus and some other security software, and perhaps even distro great how-to's (like http://www.nsa.gov/ia/_files/factsheets/Best_Practices_Datasheets.pdf).  But cyberspace reality is complex and ever changing.  

 

The idea.

 

When a user visits the login page, they get a Consent & Warning statemnet they must accept to continue (e.g.

https://www.my.af.mil) but that "I Agree" also kicks off an browser scan using BrowserCheck.  The scan's results are given to the user (typical BrowserCheck results) but also to the website's owner (as in the Business Edition).  To enter the site, the user must pass the checks or face some sorta hurdle (time delay, alternate scan process, answer manual questions about the system, etc.).  A good guy would more often than not just update the old software, rescan, and get in.  This makes that one end node more secure and thus overall improves the security of the group's data.  The site owner also gets a pretty good idea of his user's security posture. 

 

The product would be something like BrowserCheck Business Edition but with a secure backend connection with the group's authenication webserver, a simple way to feed that webSSO results.  Or perhaps the scanner is provided from that user's website (as to maintain a constant and same-named https connection between the user and the organization) but actually run by Qualsys in the cloud.  The website owners report from Browsercheck would just pretty much report %'s of before and after scans.  Maybe an advanced feature would provide an alert the webmaster of the user who always avoids updating her end node --- this may reveal a problem the user's computer, a unique device that is otherwise very secure, or a bad guy who's stupid enough not to update his own computer and thus stick out.


Anyone in Qualsys interested in discussing this idea?

 

 

 

Kevin Sweere

Air Force Research Lab

Outcomes