SlashNext launches AI internet threat protection system

SlashNext announced the company’s broad market release of the SlashNext Internet Access Protection System to protect organizations from cross platform social engineering and phishing, malware, exploits and callback attacks. The system goes beyond first generation signature-based and second generation sandbox-based technologies and deploys artificial intelligence and cognitive thinking to stop these Internet attacks targeting unsuspecting employees as their entry points.

Automatic software updates and enhanced security offered in modern browsers prevent most software exploits such as buffer overruns and privilege escalations, but social engineering and phishing attacks exploit human vulnerabilities by deceiving victims into taking actions that will breach their company and their connected client’s networks. In fact, social engineering and phishing attacks are the fastest growing security threat for organizations today, representing 43% of all Internet access threats, nearly double that of malware and viruses, according to the Verizon Data Breach Digest.  

“Social engineering and phishing attacks are becoming the prime attack vector since cyber-criminals realize unsuspecting people create the easiest way to bypass traditional anti-virus and sandbox technologies,” said Fran Howarth, Senior Analyst with Bloor Research. “A new approach to Internet threat detection is needed. SlashNext answers that need, using cognitive computing technologies that mirror human learning to enable systems to stop Internet threats targeting a variety of operating systems such as Windows, Linux, OSX, Android and iOS.”

SlashNext threat detection features include:

• A cross platform protocol analysis engine that processes gigabits of Internet bound traffic in real-time to extract a complex set of artifacts. These artifacts are essentially the telltale signs of a malicious attack


• The artifacts are further processed by a cognitive computing machine that uses massive cloud computing power to convert these features into clear Indicators of Compromise (IOCs)


• The IOCs are then handed over to hundreds of reasoning engines that behave like a team of decision-makers working together to reach a single verdict, “100% Malicious” or “Not Malicious”


• Once a decision is made, the final outcome is shared back with all the decision makers as part of a peer feedback mechanism that gives the system its unique self-learning capability.