Page 46 - Cyber Warnings

However, today’s hackers are using unrecognized, un-signatured custom codes that constantly
evolve past the latest security patches to fool and bypass legacy defenses.

Such black hat machines only need enter the organization once and from that point of entry,
they observe how to appear as authentic as real devices, servers and users.

By modelling thousands of authentic users and devices, and hiding their actions quietly among
everyday tasks instead of brute force port-scanning or head-on attacks, these automated
attackers will blend into the enterprise system, conducting lateral reconnaissance without
blowing their cover.

And before anyone can react, they can strike a fatal blow to the system in a blink of an eye once
activated.

New machine learning

However, hope is not lost in the face of evolving threats as history has shown that seemingly
insurmountable challenges can be solved through the adoption of new technology. For example,
during the Industrial Revolution, machines came to replace manual labor.

In today’s context, organized crime and nation-state groups are using automated attacks
against corporate networks, and the assaults are of such severity and speed that human
responses almost always cannot happen quickly enough.

Fortunately, thanks to recent advances in complex mathematics, organizations can now fight
back using their own machine intelligence.

A machine learning approach is synonymous with the make-up of our human immune system,
which is based on a notion of early detection and intervention to fight against viral infection.

Similarly, benign machine intelligence can act as an enterprise’s immune system, and is able to
automatically differentiate what is inherent to the body, and what is manifestly ‘anomalous’.

The enterprise immune system is able to do so because it uses advanced algorithms in a
mathematical framework to instinctively process and make sense of the torrential deluge of data
in the system, and so establishing the network’s baseline ‘pattern of life’ or what is inherent
abnormal.

Once the Enterprise Immune System comprehends a network’s ‘pattern of life’, it then makes
logical, probability-based decisions against external and insider threats at machine speed and at
scale.

46 Cyber Warnings E-Magazine October 2016 Edition
Copyright © Cyber Defense Magazine, All rights reserved worldwide