Using AI in Sandbox Earns DHS Pact

Monday, July 9, 2018 @ 01:07 PM gHale

Cyber 20/20, Inc. of Newark, Delaware, earned a contract to develop security capabilities for financial services as part of the Department of Homeland Security (DHS) Science and Technology Directorate’s (S&T) Silicon Valley Innovation Program (SVIP).

Cyber 20/20 earned its award under the SVIP Financial Services Cyber Security Active Defense solicitation, which sought solutions to cybersecurity threats faced by financial institutions. The company received $200,000 for Phase I of SVIP, which requires proof-of-concept for its proposed technology.

RELATED STORIES
Drone Sensors Focus on Security
Big Boost for Drone Forensics
Software Tool Can View Wastewater Plant Overload
Laser Sensor Can Smell Gases

“As malware evolves, so must the technology we use to defend financial services networks,” said Greg Wigton, program manager for the Next Generation Cyber Infrastructure Apex program. “Using sandboxes to predict and prevent attacks increasingly has become popular over the past decade. Cyber 20/20 plans to create a next-generation open-source sandbox that leverages machine learning to adapt to evolving threats.”

Cyber 20/20’s project — Trained Using Runtime Analysis from Cuckoo Outputs (TURACO) — expands the capabilities of Cuckoo, an open-source sandbox, to better detect and analyze malicious attacks.

A sandbox is a virtual container placed around a section of an operating system in which security analysts can force new applications to run in an isolated environment, enabling security programs to trap and analyze malware or attacks in an easily deleted setting.

TURACO combines machine learning with existing capabilities to create dynamic sandboxes that can anticipate, analyze and respond to malware attacks. TURACO also has “anti-anti-sandbox” capabilities, meaning it can defend against anti-sandbox technology, which is malware specifically designed to circumvent sandbox protections.

Companies participating in the SVIP are eligible for up to $800,000 over four phases in non-dilutive funding to adapt commercial technologies for homeland security use cases.



Leave a Reply

You must be logged in to post a comment.