Ms. Samantha Kappagoda and Dr. David K.A. Mordecai were invited to participate in a series of closed senior-level roundtable strategy and policy discussions regarding Artificial Intelligence (AI) and applications to national security. The sessions were hosted by Center for Strategic and International Studies on April 18, June 20, and July 25, 2018, at its Washington, D.C. headquarters. The discussions were held under Chatham House Rules, and supported by an ongoing CSIS study funded by Thales USA.
CSIS described the three sessions topics as follows:
The April 18, 2018 session primarily focused on establishing a conceptual framework for the application of AI to national security issues. In many national security discussions, establishing a clear and precise conceptual framework is critical in order to adequately differentiate AI applications from related but distinct technologies, e.g., unmanned systems (satellites and drones), the internet of things, robotics, and quantum technologies (i.e., computing, sensing, communication, encryption). In addition, this session sought to identify and articulate the roles for the adoption of AI, and to categorize the preexisting national security applications of AI.
The June 20, 2018 session primarily focused on issues related to the adoption of AI technologies and techniques within the Department of Defense (DoD), characterizing the current state of AI adoption in national security applications across a range of areas, e.g., logistics and intelligence, as well as a brief comparison of the state of adoption in the private sector, i.e., near term trends in adoption as well as enablers and barriers to continuing the trajectory for the DoD. Furthermore, this workshop explored incentives that might facilitate or discourage the adoption of AI technologies, identify adoption-related disruptions, and address how such disruptions might be mitigated.
The July 25, 2018 session primarily focused on issues related to the operational management of AI techniques in national security applications:
- The current state of management of AI as well as pathways to operational capability for future programs,
- Requirements presented in the use of AI in the national security mission space e.g., data ownership, software acquisition, network risk management, and user needs at the tactical to strategic level, and
- Issues of trust, verification, and reliability in AI systems, as well as the applicability of existing guidelines and practices to these systems.
Furthermore, this session introduced broader discussion regarding policy opportunities in guiding the ethical implementation and operational management of AI.
The Center for Strategic and International Studies (CSIS) is a bipartisan, nonprofit policy research organization dedicated to advancing practical ideas for addressing challenges. Founded in 1962, the stated purpose of CSIS is to define the future of national security, guided by a distinct set of values – non-partisanship, independent thought, innovative thinking, cross-disciplinary scholarship, integrity and professionalism, and talent development.