Page 25 - GCN, Feb/Mar 2018
P. 25

 The hunt for threat
Staffan Truvé
CTO and Co-Founder, Recorded Future
Machine learning is speeding the processing of cybersecurity data
are being asked to monitor, sift through and analyze vast amounts of data in order to understand the most pressing threats. They’re asked to prioritize the  xing of vulnerabilities based on highest risk to their agency. They’re asked to constantly reevaluate their threat pro le. And they’re asked to do this in the face of increasingly complex environments and sophisticated attackers.
But central to all that is getting through huge, untenable swaths of data — from both internal and external sources — to truly understand what agencies are up against. It’s impossible for humans to do this alone. Machine learning and arti cial intelligence can help.
Machine-based tools ingest data in a variety of languages and from a variety of sources, including social media feeds and the dark web, and then transform all that unstructured data into actionable threat intelligence. For example, if there’s a sharp increase in discussions about a vulnerability or threat, a machine learning- based tool can analyze those conversations and highlight what’s most important.
Similarly, machine-based threat intelligence is helping the airline industry block malware linked to an Iranian state-sponsored campaign against critical infrastructure in part by synthesizing online activity in multiple languages.
Agencies have so much data that no team of analysts could possibly rate and assess everything, and many struggle to  nd the important pieces of information. An e ective solution is one in which the simple tasks — data aggregation, comparison, labeling and contextualization — are completed by machines, leaving humans to do what only they can: make e ective, informed decisions.
Introducing the Only Universal Threat Intelligence Solution
Powered by Machine Learning Delivered in Real Time
Staffan Truvé is CTO and co-founder of Recorded Future.

   23   24   25   26   27