This weekend the latest superhero movie will open – pitting the two most popular superheroes of them all against each other: Batman versus Superman. In the film, Batman, the regular guy fighting the good fight with only his personal strength, wits, and a few gadgets, faces off against Superman, the almost limitlessly powerful alien. It is an interesting battle which I find parallels the argument between those that support user-driven, manual classification (Batman) and those that swear that only machine generated automatic classification (Superman) should ever be used.
In the business world, our “superheroes” consist of technologies that help us manage and protect data from the moment it is created to the day it is finally deleted. It would be great if we actually didn’t need superheroes to fight crime, but crime is a fact of life we cannot escape. Criminals are actively trying to steal our data and we need to protect it, be that preventing inadvertent data breaches, protecting data when shared outside our home perimeter, or safely disposing of data when it becomes a liability. So, in the battle of “Batman” versus “Superman,” who do you choose?
I think the easy answer is Superman. Super-speed and super-strength make short work of any project. In the case of classification, automation does the work of identifying and classifying data so your users are liberated from that extra bit of data housekeeping. Why lock the door when Superman is guarding the hall, right? When that report is generated from Salesforce or SAP, or when that file is downloaded from the web, TITUS automated classification is there to apply appropriate classification metadata to help ensure your data security policies are properly enforced. Unfortunately, as powerful as automated classification is, it does have a substantial weakness, and that is the inability to understand context.
Batman classification – excuse me – user-driven classification is arguably more effective than automated classification because the user creating the data understands the content and context. Users are uniquely positioned to understand the value and importance of the information they are creating. In almost every case, TITUS customers love that the user is prompted to consciously consider the value of the information they are creating and sharing as it has a huge impact on their internal culture of security – like a Gotham City Neighborhood Watch program. The biggest mark against manual classification – even though TITUS makes it easy with the one-click and favorites classification toolbars – is that the user has to take that one extra step.
So, we are left with a conflict between two heroes.
It’s at this point in our film where we see the entrance of a third hero – “Wonder Woman” – who represents suggested classification. Suggested classification blends machine classification with a request for user validation. If the machine cannot grasp the full context of the data, TITUS will present the user with its best guess, leaving it to the user to simply approve or adjust the classification to the appropriate level.
The real solution, though, is to use all three together in a kind of classification Justice League. Your organization may require that some users always manually classify data while others can mostly or even completely rely on automated classification. Likewise, some types of data are easy to automatically classify based on their originating source, location, and/or content while other data types elude the capabilities of automated identification and require manual intervention. Regardless of your preference, think of TITUS as the “Hall of Justice” for your data protection ecosystem. Multiple “heroes” provide the flexibility necessary to identify your data as your business needs require, while the classifications work in partnership with your other data security heroes (DLP, CASB, encryption, etc.), making them more effective.
So, why choose just one superhero when you can choose them all?