From: Israel Defense

AI Learning to Identify Cyber Attacks

The Israeli company Hybrid Security has developed a system called Telepath, which makes use of artificial intelligence for spotting suspicious behaviors in Web and intranet systems. The company is presently launching a second version of the system, which was installed last year in several national infrastructures in Israel in the framework of an operational pilot program.

The system includes more than ten algorithms developed by the  company, which examine the user’s surfing process and the various aspects of his behavior. The AI studies the surfers behaviors, learns what to expect and to locate the abnormalities in the framework of his network activity.

This is one of the first products in history that uses artificial intelligence in this field,” says Raviv Raz, Hybrid Security CEO. “In most of today’s cyber systems, there as a predetermined knowledgebase of signatures which the systems use to identify attacks. Our system studies on its own and becomes smarter and more precise over time, meaning that it will be able to deal with the next attack, such as attacks from Iran and China, without knowing when and how it’s about to happen.

“The US banks are under a massive cyber attack today, and the problem is so big that the US legislator passed a bill in 2012 (the FFIEC standard) that states that such technology must be installed in the systems of every financial institution. Our system is intended to deal with massive loads that don’t exist in Israel. We are beginning to work with government, communication and financial entities in the US, UK and other countries in Europe. The attacker isn‘t always the source of the attack, but could be an innocent victim himself, one whose computer was infected with a Trojan virus through which the site is attacked. Alternatively, his computer could be hacked and used to enter the targeted systems to which he has access – the “Man-In-The-Browser” attack. Sophisticated attacks such as this require locating the abnormality based on the behavioral history of the user. Man-In-The-Browser attacks have resulted in damages to European banks this year, estimated at tens of millions of Euros.”