?Mitigating Proliferation Risks Posed by Artificial Intelligence Enabled Molecular Models and Leveraging Nonproliferation Opportunities?

The U.S. Department of State's Bureau of Arms Control and Nonproliferation is seeking proposals for a project that aims to mitigate the proliferation risks posed by artificial intelligence (AI) enabled molecular models. The project will focus on identifying potential vulnerabilities and developing safeguards to counter the misuse of AI models in the chemical and biological fields. The successful applicant will be expected to develop and conduct technical evaluations and risk assessments, test established and develop new safeguards, and develop tools and methodologies to detect and counter illicit chemical and biological proliferation activities. Some key objectives of this project include: • Identifying key foreign stakeholders involved in AI enabled molecular design and production • Developing and conducting technical evaluations and risk assessments of training data sets, model architectures, computing systems, and related infrastructure used to develop AI molecular models • Testing established and developing new safeguards, mitigation measures, and best practices to secure sensitive data sets, computing systems, infrastructure, and molecular models • Developing tools, methodologies, and other resources that leverage beneficial applications of AI to detect and counter illicit chemical and biological material acquisition, development, diversion, proliferation, weaponization, and use activities The project is expected to have a duration of 24 months, and the estimated funding for this project is $4,000,000.