CSIS Futures Lab Announces Partnership with Scale AI

Read the Press Release on ScaleAI's website here.

The Center for Strategic and International Studies (CSIS) Futures Lab announces a new collaboration with Scale AI, focusing on the integration of advanced AI technologies into international security research. This partnership centers on the use of Scale AI's Donovan platform, which includes large language models (LLMs), to enhance analytical capabilities in strategic studies.

With the rapid advancement of AI technologies, notably LLMs, a wide array of fields have begun exploring their potential applications. International affairs, a sector traditionally reliant on conventional research methodologies, is now poised to benefit from these technological advancements. The CSIS Futures Lab aims to leverage AI to provide deeper insights into complex global security issues.

The collaboration will utilize Scale AI's Donovan Platform and Data Engine. This initiative is geared towards developing customized LLMs to process and analyze large-scale datasets relevant to national security challenges, especially in areas of global strategic competition. The expectation is that these AI models will identify patterns and insights not readily apparent through traditional analytical techniques.

CSIS researchers will work in conjunction with Scale AI to align the outputs of these generative AI models with national strategic objectives and international norms. Combining CSIS’s expertise in international affairs, dating back to its founding in 1962, with the technological capabilities of Scale AI, the partnership aims to advance the application of AI in understanding and addressing national security challenges.

"AI has significant potential to impact foreign policy and international security," said Dr. Ben Jensen, Senior Fellow at CSIS. "Through our partnership with Scale AI, we hope to enhance our research capabilities, enabling more comprehensive analysis of international issues and the formulation of data-driven strategic recommendations in a complex global environment."

The initial projects under this collaboration will investigate the application of LLMs in several domains, including strategic-level wargaming, domain-specific decision support and adjudication, and the development of an international relations strategy "Turing Test."