DARPA eyes artificial intelligence (AI) tools to anticipate cyber vulnerabilities at the design stage

Oct. 10, 2019
AIMEE will help anticipate emergent execution at the design stage and mitigate its propensities for cyber exploitability before deployment.

ARLINGTON, Va. – U.S. military researchers are asking industry to develop design and development tools to enable engineers anticipate and block potential cyber vulnerabilities at a system's design stage.

Officials of the U.S. Defense Advanced Research Projects Agency (DARPA) in Arlington, Va., issued a presolicitation (DARPA-PA-19-03-02) on Wednesday for the Artificial Intelligence Mitigations of Emergent Execution (AIMEE) project.

AIMEE will address the problem of anticipating, at a system’s design stage, the models of emergent execution inherent in its design, and thus mitigate its propensities for cyber exploitability before they lead to actual vulnerabilities in complete deployed systems.

Modern computing systems demonstrate strong propensity for unintended, computations and unintended, emergent programming models -- known as “weird machines” -- that enable or amplify cyber attacks. One example of this is Internet-connected toys that are exploited in denial-of-service attacks.

Related: DISA asks industry for trusted computing ways of using artificial intelligence (AI) to detect malware

Computing mechanisms built for a specific purpose and with particular models of execution sometimes can executive unintended computing tasks outside of their original specifications.

Recent research, however, strongly suggests that a system’s propensity for emergent execution can be anticipated and mitigated at the design stage, when the system’s programming abstractions and intended behaviors at a particular layer translate into more granular states and logic down in the computing stack.

The AIMEE project seeks to address the problem of anticipating cyber vulnerabilities at a system’s design stage, and mitigating these vulnerabilities in complete deployed systems.

AIMEE will explore whether a combination of recent advances in artificial intelligence (AI) techniques such as autoencoders, evolutionary programming, deep representation learning, and neural sketch learning, could detect, describe, and model the primitives of emergent execution directly in design-level prototypes.

Related: Decomposing system security to prevent cyber attacks in trusted computing architectures

Proposers should suggest how to apply AI methods to design prototype-level representations of several common computing layers known to manifest emergent execution behaviors and programming models to help anticipate these behaviors.

Strong proposals will discuss ways to generalize these case studies for use with the exemplar designs such as layered APIs, ABIs, or CPU microarchitectures, DARPA researchers say.

The project will be about 18 months long, and spend about $1 million over two phases. Companies interested should upload proposals as .zip files no later than 8 Nov. 2019 to the DARPA BAA Website at https://baa.darpa.mil.

Email questions or concerns to Sergey Bratus, the DARPA AIMEE program manager, at [email protected]. More information is online at https://www.fbo.gov/spg/ODA/DARPA/CMO/DARPA-PA-19-03-02/listing.html.

About the Author

John Keller | Editor-in-Chief

John Keller is the Editor-in-Chief, Military & Aerospace Electronics Magazine--provides extensive coverage and analysis of enabling electronics and optoelectronic technologies in military, space and commercial aviation applications. John has been a member of the Military & Aerospace Electronics staff since 1989 and chief editor since 1995.

Voice your opinion!

To join the conversation, and become an exclusive member of Military Aerospace, create an account today!