RTX BBN Technologies to develop tool that compartmentalizes software systems to prevent escalation of cyberattacks
RTX’s (NYSE: RTX) BBN Technologies was awarded a contract to support DARPA’s Compartmentalization and Privilege Management, or CPM, program. The CPM program aims to enhance cyber resilience by automatically subdividing software systems into smaller, secure compartments, preventing initial breaches from escalating into successful cyberattacks while maintaining system efficiency.
According to the U.S. Government Accountability Office, the U.S. Department of Defense has experienced more than 12,000 cyber incidents since 2015. These incidents threaten personal privacy as well as national security. The most common exploit involves a hacker gaining access to a system and then taking advantage of coding errors that allow them to escalate their system privileges to gain access to sensitive data or to take control of the system.
Under CPM, BBN is developing the Analysis and Restructuring for Containment (ARC) tool to thwart unauthorized privilege escalations and lateral movements within software systems. ARC will be engineered to automatically analyze large code bases and construct smaller, secure compartments. By applying the principle of least privilege at a sub-program level, the tool will ensure that only the minimum access necessary is granted for code to execute. This approach to software security will significantly limit the scope of potential damage in the event of a successful infiltration of the software.
“Today’s complex attack surfaces and increasingly sophisticated cyberattacks mean that even a single point of vulnerability can compromise an entire system,” said Aaron Paulos, BBN principal investigator. “Our solution will enhance the security of critical software systems while preserving performance, which is essential for maintaining operational readiness. The goal is to create compartments that isolate risks, making systems more resistant to cyberattacks.”
A key element of the program is the requirement to minimize the impact of compartmentalization on overall performance while producing secure, tight compartments. To achieve this, ARC will generate solutions that balance multiple objectives. For instance, some parts of software application will require performant access, while others might introduce significant exposure to risk. The tool will enable system administrators to selectively apply security measures to those areas deemed most critical, as a way of managing the trade-offs between performance and security.
ARC builds on several unique capabilities from BBN’s prior work in cybersecurity and software analysis. The team intends to integrate capabilities that use:
- Automated program analysis to assess and identify potential threats in software, ensuring thorough evaluation and security.
- Verifiable program restructuring to improve security and controls, including adjustments to memory and function usage.
- Automated reasoning to develop effective security solutions by exploring different options and balancing performance with risk management.
The BBN-led team includes Northwestern University, George Washington University and Kestrel Institute. Work on the program will be completed in Cambridge, Massachusetts; Evanston, Illinois; Washington, D.C.; and Palo Alto, California.
This material is based upon work supported by the United States Air Force and DARPA under contract number FA8750-23-C-B031. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not reflect the views of the United States Air Force and DARPA.
Explore AITechPark for the latest advancements in AI, IOT, Cybersecurity, AITech News, and insightful updates from industry experts!
