Leidos has won a $8.8 million contract from the Defense Advanced Research Projects Agency to develop Intelligent Generation of Tools for Security.
The INGOTS program develops automated analyses to protect mobile operating systems from high-severity chained vulnerabilities in mobile device software. This allows for timely discovery and mitigation of dangerous vulnerabilities in mobile devices, the company said.
INGOTS will also help secure mobile operating systems and applications.
“Complex system software can include a myriad of low-level vulnerabilities that can be exploited when chained together for a high severity impact,” said Artie Mabbett, leader of the Leidos Innovations Center. “Our work on INGOTS aims to automate the search through large amounts of software to identify, prioritize and mitigate vulnerability chains before they can be used in a cyber attack.”
Leidos also leverages a graph neural network to model exploit chains and uses machine learning to analyze those chains for severity and persistence. The company will also use Large Language Models to refine threat intelligence from cybersecurity threat databases.
“Our work in Trusted Mission AI really helps us in securing these systems for our warfighters because we can use it to not only confirm a patch works, but also locate other paths where the original vulnerability could possibly be utilized again,” said Ray Richards, head of LInC’s information and data sciences division. “INGOTS will take a lengthy, manual process and reduce it to a quick review that constantly keeps updating and testing to protect our troops’ communications.”