Courtesy of Morgan State University
The Office of Naval Research (ONR) has awarded Morgan State University a five-year, $9 million grant to address a critical Department of Defense (DoD) need for research in building diverse knowledge bases related to artificial intelligence (AI) and machine learning (ML), especially with respect to cybersecurity. In addition, the grant will be leveraged to elevate the University’s standing as a high research institution as ranked by the Carnegie Classification of Institutions of Higher Education. The resources will be administered through Morgan’s Division of Research and Economic Development (D-RED) to fund collaborative research conducted by the newly launched Center for Equitable Artificial Intelligence and Machine Learning Systems (CEAMLS) and the Cybersecurity Assurance & Policy (CAP) Center. The centers will study the development of formal standards and best practices to test and design new AI and ML innovations that mitigate algorithmic bias.
The funding comes as part of an overall Department of Defense $27 million investment in Historically Black Colleges and Universities (HBCUs) to conduct research in defense critical technology areas while fostering workforce diversity and entry of underrepresented minorities into science, technology, engineering, and mathematics (STEM) disciplines. The competitive awards are administered through DoD’s HBCU and Minority-Serving Institutions (MSI) Research and Education Program. Morgan was one of three HBCUs selected out of 14 proposals received and evaluated by the Military Service research offices.
“When we convened a national Blue-Ribbon Panel on STEM Research Expansion, consisting of some of the country’s leading scholars and members of the scientific research community, in 2021, and the panel recommended that Morgan focus its prominence in AI research in the formation of an interdisciplinary center to address some of the challenges of today and tomorrow, this is what we envisioned,” said David K. Wilson, president of Morgan State University. “The award from ONR further strengthens our efforts as a recognized research institution and advances our goal of achieving a Carnegie Classification of R1 (very high research activity) status over the next ten years. This is the first of many important research projects we foresee the center undertaking.”
Under the project, CEAMLS and the CAP Center will conduct research and provide education supporting the enhancement of research programs and capabilities in the critical national security functional area of trustworthy AI/ML for cybersecurity, by addressing the detection and mitigation of AI vulnerabilities due to algorithmic bias. The work will also entail the development of new formal standards and best practices in the areas of data preparation, feature engineering, model training, and deployment.
Additionally, the effort will provide education and research opportunities for Morgan students while seeking to increase the number of graduates with STEM-focused degrees to fill professional opportunities in key areas of national importance. The grant also provides for the hiring of a variety of staffers including two post-doctoral associates, three graduate students and eight undergraduates.
Designed to give computers the full responsive ability of the human mind, AI is an artificial system developed in computer software that simulates human behavior. As a discipline of AI, ML gives computers the ability to find data patterns and make predictions without being explicitly programmed. When combined, security professionals can use AI and ML to identify many types of threats, vulnerabilities, and risky behavior. Specifically, in cybersecurity, AI can reduce the time required to make critical decisions and remediate threats. The performance of AI and ML systems relies on the most accurate, effective, and efficient strategies. However, when there are biases in the development of AI and ML cybersecurity systems, the outcomes can cause potential human harm, particularly impacting underrepresented groups.
“Artificial intelligence is increasingly driving cybersecurity efforts related to the nation’s critical security posture. As technology evolves, there is a fundamental need to ensure that we employ trustworthy AI solutions in this endeavor and Morgan’s new Center for Equitable Artificial Intelligence and Machine Learning Systems will play a pivotal role in providing diverse perspectives to improve DoD applications, directly impacting our national security posture,” said Kofi Nyarko, Ph.D., director of CEAMLS and a co-principal investigator for the project. “A key part of the Center’s mission is to provide educational and experiential learning opportunities for a diverse set of students at all levels, which will facilitate the increase of STEM graduates, who will eventually provide critical contributions to the nation’s strategic challenges.”
Launched in July 2022, CEAMLS is a part of Morgan’s growing network of mission-focused research centers that serve as focal points for the University’s cross-disciplinary research programs. The center’s goal is to research how to: improve transparency and explainability in AI models; include fairness as an optimization objective; develop new tools and techniques for detecting bias and reducing the disparate impact caused; and how to develop adversarial tools to stress-test models. The Center for Cybersecurity and Policy Assurance (CAP Center) was established in 2018 to provide the electronics industry and intelligence community with knowledge, methodology, solutions, and skilled cybersecurity engineers to help prevent penetration and manipulation of the nation’s cyber physical infrastructures.
