With more breaches sensationalized in the news each week, the question, “what can we be doing better?” is sure to come up. I’ve been glad to see more being written about user behavior analytics (or UBA) over the past few months.
To quote industry analyst Rob Enderle in an article on CIO.com, “I recently attended an event where I was surprised to learn that of a number of companies that had deployed a UBA solution, 75 percent indicated they had caught a breach in progress with it. Makes you wonder how many breaches aren’t being caught in firms that haven’t deployed this technology… UBA builds a profile of each employee and if it sees an employee acting strangely it sends out an alert.”
I wonder along with Rob about the number of breaches that aren’t being caught, and they’re probably occurring far more frequently than any of us want to believe. I also agree that user behavior analysis is definitely another valuable piece of the security puzzle – in fact, it’s one of the only things that can help detect rogue insiders, or outsiders that have compromised the credentials or systems of insiders.
However – and though it may seem obvious – in order to perform user behavior analysis, you have to track enough behavior to have something meaningful to analyze. Unfortunately, very little is actually recorded about what employees do with corporate data, much less analyzed.
From a Ponemon Institute survey sponsored by Varonis, only 22% of employees say their organization can tell them what happened to lost data, files, or emails. It’s not just files and emails missing that they don’t know about – they don’t know who is opening them, sending them, creating them, or changing them, or when or which devices they’re using.
Conducting user behavior analysis without looking at how and when employees access data is like credit card companies trying to do fraud protection without looking at credit card transactions. It’s absurd. If you want to protect data, you have to watch how it’s used normally every day, baseline normal behavior for each user, and then try to spot the abnormal.
That’s not to say you can’t do a lot of good by tracking VPN login times, IP addresses, bandwidth consumption, etc. – the more data points you have, the more complete your baseline. Just don’t forget the essential ingredients: who is actually using all that valuable data and what they’re doing with it.