Nichols Martin
BAE Systems has received a potential $5.2 million contract from the Defense Advanced Research Projects Agency to develop automated tools for the analysis and detection of cybersecurity risks to large-scale enterprise networks.
The company said Tuesday it will help DARPA create, evaluate and demonstrate platforms that can help organizations defend against sophisticated threats as part of the agency’s Cyber-Hunting at Scale (CHASE) program.
BAE is developing a system that uses machine learning and cyber attack models designed to recognize potentially concealed anomalies.
“Our technology aims to alleviate resource constraints to actively hunt for cyber threats that evade security measures,” said Anne Taylor, product line director for BAE’s cyber technology group.
“One of the major challenges modern network defenders have is that cyber attackers can buy copies of common network defense tools, and ensure their attacks are not detected before unleashing their attacks against operational networks. Because of this, it is almost impossible for network defense tools to consistently prevent well-resourced cyber attack teams from penetrating enterprise networks,” commented Sam Hamilton, a chief scientist in BAE’s cyber technology group, speaking with GovCon Daily Wednesday.
“CHASE combats this issue by introducing a dynamic element to the attack defense process. Even if attackers have access to the CHASE defensive toolset they cannot effectively try out their attacks first, as the dynamic mechanisms (machine learning & dynamic sensor deployment) in CHASE ensure the deployed enterprise defense will act differently than the privately deployed test version an attacker might use for testing out attacks before launching them,” Hamilton added.
The company intends for its platform to aid government, military and commercial network security efforts.
Digital Operatives; Ruslan Salakhutdinov of Carnegie Mellon University; and the Farshad Khorrami-led team of New York University serve as BAE’s partners on the CHASE program.
