This last few weeks have seen, yet again, some pretty significant hacks (namely the Evernote hack). Large amounts of user data, including passwords, were released into the wild. The situation could have been worse in the Evernote case, but at least the passwords were salted and hashed. Evernote's response, has been to perform a mass password reset on it's user base, in a proactive damage limitation style exercise and no doubt several internal streams of investigation will be looking for the who, what and why behind the attack.
Security is Reactive
I've blogged on the topic of reactionary security on a few occasions recently ("Protection Without Detection", "Preventative -v- Detective Security") and it seems the approach is still pretty much the default (Or perhaps security is only really questioned and tested after an event?). There are of course, lots of 'pro-active' components to security. Hashing a password could be seen as one for example, but many of these activities are often small tactical steps at the implementation level, not the strategic level. Audit is obviously detective, with an audit response proactive in some sense, but really only proactive to get you back to the status-quo of reactive. Big data for security, ("Security Analytics: Hype or Huge?", "Big Security Data to Big Security Intelligence") is another example in my mind or purely reactive security. The big data promise is based entirely on scale and speed. Scale obviously (the word big might help there), with regards to aggregating and correlating multiple data sources and speed, for trying to develop queries and analytic steps to identify root causes, patterns and so on. Longer term, the results of the big data analytical steps could of course be proactive in nature.
A Different Approach
In my mind a different approach is needed. I'm not advocating what that approach should be, but the panacea would of course be to get security so embedded, it seamlessly integrates with revenue generating business focused practices within an organisation. The gap between security and convenience, needs to be minimized to as close to zero as possible. Security really needs moving up the organisational food chain, away from the bigger, faster, shiny implementation level approach, which will constantly chase (and lose) an attackers tail, to be a default stance with all business related policy decisions. This is difficult of course, but in the longer term will help move away from a reactionary standpoint to something resembling security by default.