Today I shared the newly released Umbrella Security Graph with my peers at the Kaspersky Threatpost Security Analyst Summit. This research tool speeds and improves the discovery, tracking and – potentially –  takedowns of complex malware and bot networks. For example, our team recently used the Umbrella Security Graph to discover different attributes, locations, and domains connected with the ‘Red October’ attacks.  That’s covered in more detail in today’s Dark Reading article about the Umbrella Security Graph.


OpenDNS and Umbrella Security Labs have been using the technology we are exposing today to keep Umbrella customers safer with predictive analysis that’s based on our data mining and classification methods.

Shortly after I joined OpenDNS last year I blogged about the need for security vendors to evolve from a strategy of ‘collect and react’ to a foundation of ‘real-time adapt’. It’s the key to ensuring that security vendors move from being a step behind cyber criminals to a step ahead of the pace of technological change. Part of achieving this vision involves effectively harnessing and analyzing Big Data so that we can predict unknown threats, rather than simply block known threats.

The Umbrella Security Graph enables our researchers to access a global view of Internet connections and traffic patterns and apply sophisticated analytics and scoring capabilities. Combined with machine learning, graph theory and related algorithms, the Umbrella Security Graph allows us to deliver predictive Internet security protection to customers. It’s just the beginning of our move into predictive security intelligence. We look forward to seeing how partners might contribute additional data scoring techniques, data mining capabilities and visual graphs to help us further transform Internet security.

We will be announcing more details on how to get access to the Umbrella Security Graph in early March. Please follow us on Twitter to stay up-to-date @thinkumbrella.

This post is categorized in: