Key Takeaways From the NSCAI’s Final Report on Artificial Intelligence
Logan Jones, General Manager and President, and David Mazar, Director of Strategy and Market Development, at SGS | April 26, 2021
The National Security Commission on Artificial Intelligence (NSCAI) recently published its Final Report — a 700+ page piece that is a culmination of two years of work resulting from a bipartisan congressional directive that asked the commission to examine AI in regards to its potential application for both the US government and military’s use. It is perhaps the most consequential report to be published on US national security since the 9/11 commission.
The NSCAI assesses the current state of the world in relation to AI while working on a strategy to expand the National Security Innovation Base (NSIB), the various organizations bringing novel technologies and innovation to the US government for national security missions. It further states the NSIB should be used as “a pipeline to resource, prioritize, acquire and iterate [AI] capabilities critical to sustain the competitive advantage.” The consequences of this report and how it will shape the US military’s strategy cannot be understated.
The potential AI has and the impact it has already made globally has elevated it to a level of national importance. America’s technological predominance, which is “the backbone of its economic and military power,” is under threat unless the US takes a leading role in AI development and implementation. Thus, the NSCAI’s findings serve to inform the US government’s strategy in how it will adopt and integrate AI as a core technology that can “comprehensively address the national security and defense needs of the United States.”
At SGS, we’d like to provide our unique perspective and highlight key takeaways around the national security and defense market issues touched upon. We would be remiss if we didn’t share the inspiration taken from this report given that former Deputy Secretary of Defense Bob Work is the Chairman of our board and was appointed to serve as the Vice-Chair of the NSCAI. As a company bringing solutions to support the adoption and expansion of AI within the national security and government market, we fully back this report, its recommendations, and the potential it has to impact and advance the applications of AI technology to support the US government.
The NSCAI minces no words about how imperative it is to find success in AI adoption. Leveraging AI is a must for national security as global threats and advances in AI abroad threaten the US’s position as a government and military leader. If one of the US adversaries, who already have some technological advantages over the US government today, becomes the clear leader, then their values will become the defacto standard, threatening western values.
This is why the NSCAI has set a clear goal – by 2025, the US must be AI-ready.
Given the report’s findings and recommendations, we’d like to share our perspective from an AI startup focused on the federal and national security market to add to what is an incredibly important dialogue that in many ways will shape the course of history moving forward.
1. Building an AI enabled enterprise starts with the foundation
“By 2025, the foundations for widespread integration of AI across the DoD must be in place.” – The National Security Commission on Artificial Intelligence’s Final Report (p. 61)
Just like a house needs a foundation, AI requires a foundation so it can be properly adopted, applied, and integrated within any organization or department. This is true for the application of AI to any problem or use case and applies to a larger extent with an organization like the US government. In the US’ case, artificial intelligence must be approached holistically within the defense industry in order to be successful.
Whenever any organization or department, private or public, wants to leverage AI, the key question is: “Are you currently collecting, or are you able to collect data associated with the problem you are interested in solving?” Establishing data readiness may not be the most exciting step, but it’s a necessary one that will serve as a starting point for achieving AI implementation success.
Answering this question lies beyond just data and technology. It requires educating the workforce who engages with the technology and shifting acquisition approaches to align with this new era of AI integration. Establishing data readiness, properly aligning AI acquisition and adoption, and approaching the process holistically will result in a strong foundation to build upon and will place the military in a prime position to leverage AI in the best way possible.
2. Leveraging commercially derived AI will drive returns and quicker deployment
“DoD should embrace proven commercial AI applications and incentivize their use to generate labor and cost savings, speed administrative actions, and inform decision-making.” – NSCAI Final Report (p. 65)
In order to succeed with AI, the DoD must rely on commercially derived AI solutions that bring benefits, knowledge, and lessons from success in the private field. Embracing commercially derived AI solutions will require adopting a modern approach that better balances risks and outcomes and eschews stereotypical bureaucratic processes that will only hinder future partnerships with technology firms as well as any critical efforts to expand the National Security Innovation Base.
The old approach has deterred companies from working with the DoD as it is economically irrational for many startups to even consider it, given the lengthy sales cycles, financial requirements, and overwrought regulations. As a response, the NSCAI recommends that the “DoD should create opportunities for bottom-up identification of AI use cases by incentivizing DoD organizations to deploy proven commercial applications tailored to their specific mission needs.” (p. 307). By taking this bottom-up approach and calling out the need to deploy commercial applications, the DoD can evade issues that historically plagued them and complicated work with past commercial companies and partners.
However, we should note that the DoD should identify commercial partners with a government-ready solution that has tailored its products and process specifically for government adoption, not just a ported-over solution. It’s critical that the solution is focused on a government AI solution incorporating benefits and intelligence from past commercial success and building on them with the unique needs of the government customer they are supporting. Ultimately, the US is best served by an AI partner rather than a company who will treat the government like just another customer.
To specifically address and reach this goal, the report has already identified the Joint Artificial Intelligence Center as the DoD’s AI accelerator to speed up and scale AI adoption across use cases and serve as the technological bridge between the DoD and other government agencies. This will allow the US government to have the modern approach to AI (and other tech) adoption needed to ensure it maintains its position as the global technology leader.
3. The US must align the acquisition approach to the technology being acquired
“The Department must have an integrated approach to AI and other emerging technologies that ensures the U.S. military can continuously identify, source, field, and update capabilities faster than our competitors.” – NSCAI Final Report (p. 302)
As the NSCAI rightly calls out, the current acquisition approach the DoD employs requires long cycles to deploy resources, a process that works for legacy and traditional hardware acquisitions but not for AI acquisitions or partners. To find success in this initiative, the DoD and US government needs to update and rehaul their AI integration, adoption strategy, and framework to account for all aspects of this initiative to bring forward bold, realistic, and compelling outcomes. The military can’t afford to rely on 18-24 month timelines to acquire technology because the technology quickly becomes outdated and superseded by newer, more innovative developments–further holding the military back due to a reliance on old technology.
An essential component of aligning the acquisition approach also requires training and upskilling the department’s workforce to know how to scope software projects, integrate solutions, and buy AI technology.
To achieve this, the report recommends creating a digital service academy and civilian national reserve to grow tech talent with the same seriousness as military officers. Further, it recommends the DoD establish integrated AI delivery teams at each Combatant Commands and work with other allies and partners to create AI interoperability and further foster adoption of the technology.
Properly aligning acquisition to AI will require an entirely new framework and infrastructure; innovation is needed across every step of incorporating AI into the DoD, from training to acquisition to adoption and application.
4. Fund technology development and capability expansion, manage risk
“The Department should focus on four actions…4) Update the budget and oversight processes. DoD’s resource allocation process is nearly identical to what was put in place in 1961. It is incompatible with AI and other digital technologies.” – NSCAI Final Report (p. 66)
Moving forward, the US will need to fund deployment based on faster technology development cycles and dynamic threat environments. The NSCAI report has prioritized this initiative by advocating for funds to transition technology adoption and shift the culture to become less risk-averse. This will allow the US to not only enable future innovation but drive the adoption necessary to scale AI.
Any new technology development and expansion requires investment and appropriate risk management — this approach will require some acceptance of risk as well as a cultural shift allowing for a “fail fast and learn” mentality, rewarding small failures and pivots as well as large, scalable successes.
The report also recommends taking a portfolio management approach to budget and requirements. This would be a significant step in the right direction and would be a positive example of the department adopting a commercial best practice. To find the same success within the DoD, it will require a partnership between all parties in any new technology or resource deployment.
The same traditional approach to technology that has led to the glacial pace of digitalization must also be amended in regards to risk management. Fortunately, the NSCAI’s report outlines key components that align with this thinking. Funds will be allocated to support this initiative, and the report recommends the DoD “[r]etire legacy systems ill-equipped to compete in AI-enabled warfare” (p. 69) while also recommending them to evaluate AI alternatives before funding any other major defense acquisition.
Crucially, the NSCAI makes several recommendations on funding key efforts associated with AI development, making a clear case for material investment. A few of the recommendations (among many investments outlined) include:
- Direct the Secretary of Defense to establish a $200M+ dedicated AI fund
- Increase investments in AI research & development by $8B
- Establish a $100M fund to provide matching contributions to service and agency efforts based on estimated ROI that results from the use of AI in predictive analytics for maintenance and supply chain optimization for all classes of supply equipment and parts
- Establish a $100M fund under the management of the JAIC to accelerate procurement and integration of commercial AI solutions for business applications
5. The US must expand the defense industrial base
“The DoD should formulate its investment strategy…take into account industry’s comparative advantage in available R&D capital and include a consistent and transparent approach to messaging defense technology priorities to build and broaden the Industrial Base.” (p. 319)
The US government needs to incentivize its departments to foster a culture of innovation and AI adoption. In order to do that, the NSCAI report recommends the DoD develop a Technology Annex to the National Defense which “should identify emerging technologies and applications that are critical to enabling specific capabilities for solving the operational challenges outlined in the NDS (National Defense Strategy).” (p. 319). This will allow the DoD to ensure any new partners it works with are directly addressing identified needs and priorities.
By expanding the defense industrial base, the DoD can prime its departments and organizations to adopt any new technology that could serve its priorities and needs, starting with AI. This also means the government should expand beyond working with the same traditional players to enable new capabilities and innovation. This requires taking the appropriate risk and partnering with new companies that have promising technology that will foster new opportunities. Ultimately, this approach will create a broader pool of capable companies with the desire and ability to support the DoD in the AI market space.
This would allow AI companies to identify infrastructure requirements, establish roadmaps, and prioritize necessities while considering technical capabilities that can adapt flexibly in a way the DoD never could before.
Ultimately, the support and infrastructure the Technology Annex would provide could fast track adoption and acquisition of any technology by scaling and leveraging DoD’s R&D resources to help them meet their objectives.
The DoD and US must make AI a priority, today
As a leader in the AI industry, we stand ready to advance one of the most critical missions of our country and allies. The publication of the NSCAI report is significant at a time when the global race for AI dominance is rapidly evolving. Its recommendations are a major step forward in the right direction and timely given the President’s 2022 budgetary recommendations for the DoD.
Leveraging commercially-derived AI is absolutely necessary for faster deployment and success. At SGS, we have roots in the commercial sector with proven solutions that have matured over many years working with leading global commercial companies. Our board and leadership team bring unique insights and direction stemming from decades of national security experience, and our partnerships with key companies have enabled us to understand how to best work with the government. Most importantly, we have tailored our business structure, AI solutions, and products specifically to work with the government and to facilitate adoption and acquisition to the speed outlined by the NSCAI’s report.
Across people, processes, and funding, we must be willing to take measured risks to advance our role and capability as leaders in the AI field. It is the best way for the US to maintain its position as the global technology and military leader.t