CSET Research Fellow Margarita Konaev assessed the importance of building trust in AI systems for the DOD's successful use of AI in the battlefield. See FedScoop's interview with her and readout of her research below.

National security leaders view AI as a priority technology for defending the United States. This two-part analysis is intended to help policymakers better understand the scope and implications of U.S. military investment in autonomy and AI. It focuses on the range of autonomous and AI-enabled technologies the Pentagon is developing, the military capabilities these applications promise to deliver, and the impact that such advances could have on key strategic issues.

This brief examines how the Pentagon’s investments in autonomy and AI may affect its military capabilities and strategic interests. It proposes that DOD invest in improving its understanding of trust in human-machine teams and leverage existing AI technologies to enhance military readiness and endurance. In the long term, investments in reliable, trustworthy, and resilient AI systems are critical for ensuring sustained military, technological, and strategic advantages.

The Pentagon has a wide range of research and development programs using autonomy and AI in unmanned vehicles and systems, information processing, decision support, targeting functions, and other areas. This policy brief delves into the details of DOD’s science and technology program to assess trends in funding, key areas of focus, and gaps in investment that could stymie the development and fielding of AI systems in operational settings.

Today’s research and development investments will set the course for artificial intelligence in national security in the coming years. This Executive Summary presents key findings and recommendations from CSET’s two-part analysis of U.S. military investments in autonomy and AI, including our assessment of DOD’s research priorities, trends and gaps, as well as ways to ensure U.S. military leadership in AI in the short and the long term.

Over the last decade, Moscow has boosted funding of universities and implemented reforms in order to make Russia a global leader in AI. As part of that effort, Russian researchers have expanded their English-language publication output, a key—if imperfect—measure of the country’s innovation and impact. Between 2010 and 2018, the number of English-language publications by Russian scientists in AI-related fields increased six-fold.

CSET hosted WestExec Advisors' Michèle Flournoy and Gabrielle Chefitz, together with Avril Haines, for a discussion of their new report outlining how the Department of Defense can adapt its test, evaluation, validation and verification (TEVV) infrastructure for artificial intelligence. The authors were joined by Johns Hopkins Applied Physics Laboratory's Ashley Llorens, the Joint Artificial Intelligence Center's Dr. Jane Pinelis, and moderator Richard Danzig.

China’s surge in artificial intelligence development has been fueled, in large part, by advances in computer vision, the AI subdomain that makes powerful facial recognition technologies possible. This data brief compares U.S. and Chinese computer vision patent data to illustrate the different approaches each country takes to AI development.

Axios Future highlighted a series of one-pagers issued by CSET providing AI policy recommendations for the next presidential administration to consider and implement. The full piece from Axios can be found below.

Since 2016, China has engaged in a nationwide effort to "merge" AI and neuroscience research as a major part of its next-generation AI development program. This report explores China’s AI-brain program — identifying key players and organizations and recommending the creation of an open source S&T monitoring capability within the U.S. government.