It’s better suited to SC in AI in particular
Stanley-Lockman 21 Zoe Stanley-Lockman, analyst researching military innovation, emerging technologies, and defense cooperation, Center for Security and Emerging Technology, “Military AI Cooperation Toolbox: Modernizing Defense Science and Technology Partnerships for the Digital Age,” CSET Issue Brief, August 2021https://cset.georgetown.edu/wp-content/uploads/CSET-Military-AI-Cooperation-Toolbox.pdf /GoGreen!
Bilateral Military AI Agreements and Cooperative S&T Blueprints
Whereas the two tools above consist of structured opportunities for military AI cooperation, a host of other initiatives offer additional useful ways forward on military AI cooperation. The selection here focuses on bilateral AI agreements that create linkages to foreign national security innovation bases, as well as pathways that do not yet orient toward AI but could with only minor adjustments.
Bilateral Military AI Cooperation Agreements
DOD already has many project arrangements that create ad hoc cooperation opportunities with defense technology communities in other countries, but only a small number of recent initiatives target new forms of military AI cooperation. 55 This section focuses on two such formats, one with Singapore and the other with the U.K. Military AI collaboration is already taking place. The JAIC technology collaboration with the Singapore Defence Science and Technology Agency, agreed in June 2019, is different from other AI-related project arrangements in that it names a specific area of cooperation, and is not necessarily service-specific.56 This collaboration focuses on the mutual operational interest of humanitarian assistance and disaster relief, a relatively noncontroversial area.57 This echoes calls from several analysts to begin international military AI cooperation with lower stakes and the possibility to scale into larger initiatives.58
The centralization of AI efforts in the JAIC may portend an eventual hub-and-spokes model for cooperation. This is because its location and reporting structure in the Office of the Secretary of Defense (OSD) could promote coordination with the military departments to either work across combatant commands or cut across operational domains. Further, the coordinating role played by the JAIC could make the cooperation more visible and easier to assess as a blueprint for other countries. On this note, it could be seen as a pilot not just for U.S.-Singaporean collaboration, but also as an inspiration for similar efforts with partners in the Indo-Pacific region, which are under-represented in the toolbox here.
With the U.K., the Bilateral Academic Research Initiative (BARI), launched in 2018, focuses on creating academic collaborations. More specifically, this collaboration could take the form of grants and fellowships on AI research. With the inaugural pilot program selected to focus on human-machine teaming, DOD and the U.K. Ministry of Defence will respectively sponsor up to $3 million and £1.5 million for the multidisciplinary, academic team to conduct the “high-risk basic research as a bilateral academic collaboration.”59 Notably, while the allies with the same level of intelligence access as the U.K. are numbered, the broader academic and scientific focus of the BARI pilot could be replicable for basic and applied research with many more countries willing and able to co-fund these ventures.
The flexibility of these forms of collaboration is overall a strength. That said, for both examples here, the military AI collaboration projects fit into existing, overarching frameworks of cooperation, as declared by the heads of state or defense minister counterparts.60 Without these broader frameworks, similar working-level relationships may be seen as ad hoc.
Cooperative S&T Blueprints
Whereas the above bilateral cooperation agreements are already oriented toward AI, the defense technology cooperation tools discussed below could be integrated into the toolbox for AI, but have not yet done so.
One example of such S&T blueprints is the Allied Prototyping Initiative. Launched in 2019 by the Office of the Under Secretary of Defense for Research and Engineering (OUSD/R&E), the Initiative funds a small number of cooperative operational prototypes related to the top DOD modernization priorities. Under the Initiative, the United States can identify transition pathways for operational capabilities in the five- to seven-year horizon, including future codevelopment with the U.S. national security innovation ecosystem.61 So far, the program has funded prototypes for highspeed propulsion technologies with Norway and hypersonic vehicles with Australia.62 But given its focus on DOD technology modernization priorities, the Initiative can also oversee prototyping related to AI and machine learning, autonomy, microelectronics, and fully networked command, control, and communications.63
Building on existing international cooperation agreements such as IT&E agreements, future Allied Prototyping Initiative project arrangements can help further identify the extent of TEVV and technology transition funding focused on AI-enabled and autonomous systems. In fact, it may be better suited for this task than the longer-standing, counterpart program, Foreign Comparative Testing.
Another small, regional agreement that can serve as a blueprint for AI cooperation is the Polar Research, Development, Testing, and Evaluation arrangement between the United States, Canada, Denmark, Finland, New Zealand, Norway, and Sweden.64 Just as this agreement is based on a “shared desire for strong cooperative relationships that preserve safe, stable and secure Polar regions,” other minilateral arrangements—such as the Quad—could use this as a blueprint to align R&D and TEVV priorities with a free, open, secure, and prosperous Indo-Pacific.65