A coordinated AI effort between the US and its allies improves interoperability and increases security
Imbrie et al. ‘20 (Andrew Imbrie, Senior Fellow at Georgetown's Center for Security and Emerging Technology; Ryan Fedasiuk, Research Analyst at Georgetown's Center for Security and Emerging Technology; Catherine Aiken, Director of Data Science and Research at Georgetown's Center for Security and Emerging Technology; Tarun Chhabra, nonresident fellow with the Center for Security, Strategy, and Technology at the Brookings Institution; Husanjot Chahal, Research Analyst at Georgetown University's Center for Security and Emerging Technology; February 2022; “HOW THE UNITED STATES AND ITS ALLIES CAN DELIVER A DEMOCRATIC WAY OF AI”; CSET; https://cset.georgetown.edu/publication/agile-alliances/)-amc
The United States and its allies should also consider wargaming and table-top exercises to explore how sharing selected government data sets could shore up de- fenses against counter-AI techniques and other efforts to exploit the vulnerabilities of AI systems. Specifically, they should explore how sharing militarily relevant data sets and certain AI algorithms could help allied countries better test system robust- ness, expose mutual vulnerabilities, accelerate development of countermeasures, and establish common standards for testing, verification, and validation.68 The United States and its allies should define common standards for the level of robustness required for a given operation. Common defense planning and capabil- ity development in NATO and the EU should give priority to investments in AI safety and security, as well as common verification procedures for AI-enabled, safety-crit- ical systems. To ensure allies store and process data homogeneously, the United States and its allies should launch an accelerator fund for cloud computing. The United States and its allies could use this fund to more efficiently procure commercial cloud com- puting technology. The United States, United Kingdom, and Canada, for example, could agree to bid out a bulk purchase of cloud compute from major technology companies and distribute access to compute in the form of credits and publicly funded research. This initiative would ensure that democratic nations benefit from techniques in machine learning that require fewer inputs of real-world data but greater computational power to run simulations and self-play methods. Representa- tives from Japan, South Korea, the Czech Republic, Lithuania, and the EU each cited increased computing as an AI R&D priority, suggesting an area for aligning focus among allies. Parallel to this effort, the United States and its allies should launch a software development initiative. This initiative could take a page out of the U.S. Air Force’s Kessel Run project by pairing government-led teams with software developers from allied countries. Multinational teams could work together to build capabilities in agile software used in military systems that are part of joint exercises. Allies could also use AI to automatically create “translators” between systems and user interfac- es that are not yet fully interoperable. The United States and its NATO allies should consider partnering with existing frameworks like the AI4EU artificial intelligence test bed, which pools compute and data among EU countries.69 Optimal Partners: Canada, Australia, United Kingdom, Germany, Italy, and Japan Multilateral Fora: Five Eyes, NATO, NATO-EU (AI4EU) test bed partnership, U.S.-Japan-South Korea Trilateral Defense Cooperation, National Technology and Industrial Base (Australia, Canada, the United Kingdom, and the United States) Criteria for Partnership: The United States should improve technical interopera- bility with the countries that receive the most attention in U.S. global security opera- tions, interact the most with U.S. forces, and express the most concern about disjoint- ed technical requirements and capabilities.