A New Approach to Security Detection
To resolve the problems described above, the SlashNext Active Cyber Defense System and iPoint Networks uses neither signatures nor sandboxes. Instead, artificial intelligence algorithms extract a complex set of “features” from the entire infection sequence to determine if a flow is malicious or not. Because these features are not solely dependent on the actual object (or payload), we inherently do not
suffer from the same limitations as signature or sandbox based systems – making our detection
capabilities substantially more difficult to evade. Additionally, because the majority of our detection
logic resides in the cloud and we have strict control of every network that routes traffic to our
cloud, there is no way for a cyber-criminal to devise a test suite that checks whether our cloud will
discover his malicious code or not.