By Dr. Simon Moores, Security Futurist And Risk Consultant
If you think about some of the numbers around cybersecurity, it becomes quite scary:
Threat diversity can only increase because of the growing number of objects, which are available for adversaries themselves to exploit, particularly with the arrival of the Internet of Things.
One that is getting a lot of interest from companies, such as IBM and Google, and others is Artificial Intelligence (AI). We have the example of the world’s greatest grandmaster, Garry Kasparov, beaten by IBM’s deep blue computer and, much more recently, Google’s deep mind achieving a victory of an even more complex game called Go, against a grandmaster in Korea.
What we are starting to see is the evolution of something called deep learning and machine learning which we can start to think of as a solution to be used in the information security space.
When applied to games such as Breakout, what becomes interesting over a period of time is the machine itself starts to work out ways of winning at the game, so in a way the machine starts to work recursively and this is a characteristic of machine learning. The machine itself is not constrained by human experience and it is not constrained by expectations.
As the complexity of the attacks being directed against organizations moves beyond the capability of human operators to deal with, so in the future we will start to see more and more examples of AI starting to make an impact in the information security space alongside analytics, predictive analytics as well and much more.