UC security: When the shoe won't fit, compress the foot
- 15 April, 2009 05:06
This is a pattern that keeps repeating in many different areas: the security paradigm looked good until a technology comes along, changes the assumptions and reveals the inadequacy of the model.
So what do most companies do? They delay adoption of the disruptive technology and try to change it in a way that makes the inadequate security paradigm look less inadequate. If the paradigm is broken, pretend it isn't and try to shoehorn the new technology into a broken model.
Take unified messaging, a key component of UC. Different types of messaging, such as voice mail, fax, e-mail, which were once held in separate locations are merged into a single mailbox. The security, however, was location-centric.
It treated messages differently based on where they were stored. So stuff in the PBX (voice mail) required four-digit pin to access and was deleted after one month with no archiving. Stuff in the fax server was kept forever and required directory authentication.
E-mail had yet another set of policies. The implementation of these security measures was dependent upon location: if it's in the PBX, this is how it works. Unified messaging comes along and messes up this model. So what do many companies do?
They create multiple in-boxes with different retention policies for each user, to store the different types of messages. When lacking a security model to deal with unified in-box, don't fix the model - un-unify the in-box!
The correct response is to realize that the security policy is based on a mistaken requirement/assumption of "single set of retention policies per storage location". This assumption should be replaced with "retention policies depend on the content of a message, not its location".
Unified messaging is responsible for revealing the flaw in the assumption, but is not the cause of the flaw. Because it is a new technology and because it is disruptive there is a tendency to assume that what worked before should work now and therefore the new technology is at fault - let's fix it!
Information security today seems to be suffering from many such flawed assumptions.
Many of the old, comfortable paradigms are falling prey to disruptive technologies, leaving security professionals wondering how to respond. Perimeter security, IP-address ACLs, static segmentation, signature-based intrusion-prevention systems and antivirus are all failing models in the face of mobility, UC, virtualization, polymorphic threats and so on.
When a disruptive technology is revealing a flaw in your security assumptions, don't shoot the messenger.
Re-evaluate your assumptions instead of trying to fit the new thing into an antiquated paradigm.