National Security Commission on Artificial Intelligence Recommendations for Defense AI Investment

Key Takeaways

• The NSCAI recommends that Congress provide $14B in funding for AI-related work at DOD.
• NSCAI recommendations have in recent years made it into legislation like the NDAA.
• Recommended funding is centered in the Fourth Estate.
• Nearly all of the recommended funding is contractor-addressable.

The National Security Commission on Artificial Intelligence (NSCAI) recently published a 756-page report detailing an aggressive national strategy to bolster U.S. investment in artificial intelligence. These types of reports typically generate only marginal interest because neither the Department of Defense nor the U.S. Congress are bound to implement the recommendations contained in them. This report is different, however, because Congress created the NSCAI in 2018 and later incorporated a number of the Commission’s earlier recommendations into the National Defense Authorization Act (NDAA) for Fiscal Year 2021. Assuming Congress elects to do the same for the FY 2022 NDAA, it could pay to know what types of investment the NSCAI’s final report recommends for DOD and which organizations may receive funding related to artificial intelligence in the years to come.

Arguing that the United States is falling behind peer great powers in artificial intelligence investment, the NSCAI recommends that Congress funds just over $14B in AI-related programs/initiatives at the DOD. The specific programs/initiatives called out in the report, as well as the organizations which should oversee those programs are as follows:

As we can see, the vast majority of the recommended funding would go to efforts that are contractor-addressable. The lion’s share of recommended funding ($8B) would go to Research and Development work that only a certain type and number of contractors are prepared to pursue. Hundreds of millions of dollars, on the other hand, would go to work that a larger number of contractors would find more accessible, including systems engineering to prepare data for AI capabilities and enterprise architecture. Large sums would also go to procuring commercial capabilities for rapid adoption by the DOD. DOD would undoubtedly purchase some of these as prototypes via Other Transaction Authority agreements, but others may be procured through standard acquisition channels.

One other observation worth making is that all of the recommended funding is slated for Fourth Estate organizations despite the fact that the Military Departments (MILDEPS) are also doing a growing amount of AI-related work. It remains to be seen if Congress or the Office of the Secretary of Defense would redirect some of the recommended funding to the MILDEPS in order to meet requirements developing there. Regardless of how the funding is ultimately disbursed the fact remains that the NSCAI has set an aggressive agenda at Congress’ request that may turn into significant business opportunity for companies working in the AI space.