Developing A Robust Software Assurance Tools for Cyber Security

BlueRISC proposes a fundamental software assurance model upon which an automated tool will be built for inferring as to exploitability as well as malicious nature of the software under investigation. The proposed model enables a passive characterization of software that is able to drive an active testing framework complemented with machine learning. This reduces false positive and false negative of vulnerable codes as well as packed, malicious codes in heterogeneous computing architectures. The proposed tool is able to operate at the binary-level, increasingly breadth of coverage as well as scope, and removes the reliance on the availability of source-code.