Toward an AI-Ready Department of Defense: Building An Open Defense Ecosystem

By Logan Jones
June 21, 2021
Feature

Summary

If the Department of Defense (DoD) is to fulfill the recommendations of the final report from the National Security Commission on Artificial Intelligence (NSCAI), it must become a fully AI-ready enterprise by 2025. This goal is at risk without a change in its traditional approach to data and acquisition.

Toward an AI-Ready Department of Defense: Building An Open Defense Ecosystem

Recent steps forward, such as the “Creating Data Advantage” memo signed by Deputy Secretary of Defense Kathleen H. Hicks, are encouraging, but still only a start. While that memo describes “maximized data sharing and rights for data use” as a goal for the Pentagon, and lays out a transformational timeline, it remains to be seen how well and how completely that transformation will be implemented.

Meanwhile, the mandate for such a transformation is both obvious and irrefutable. Foreign adversaries already threaten the United States’ slim lead in AI, and over time, as the defense landscape changes from hardware-focused warfare to software-focused warfare, AI will play an increasingly important role--in fact, a decisive role. Clearly, we must accelerate and redouble our efforts in AI development and deployment to prepare for that inevitable time.

In SparkCognition Government Systems’ (SGS) recent blog post, we highlight some of the NSCAI final report’s key takeaways. Following through on those takeaways will, however, require substantial changes in the way the Department of Defense conducts business with both traditional and nontraditional contractors. In particular, we would like to call attention to the acquisition process, which at present is poorly suited to the goal of making the federal government an AI-ready enterprise in four short years.

One primary problem with the current acquisition process: its long cycle of procurement and resource deployment. Such an approach may be adequate for hardware acquisition because hardware, as a class of defense asset, changes relatively slowly. In software, however, and in particular in AI, the pace of change is rapid. Software requires a far more agile acquisition cycle to ensure our warfighters are ready for any threat.

If the Pentagon continues to use the same acquisition strategy for both hardware and software, it will be too slow to take full advantage of the best available technology. Instead of leveraging AI to respond swiftly and effectively to unpredictable changes and challenges, our national defense will be hobbled by last-generation solutions of reduced efficacy. The overall outcome will be that foreign adversaries are empowered to create and extend a critical lead in AI--outcompeting the United States in an area where we should logically be the unquestioned leader.

The current acquisition process has gradually developed out of the long history of relationships between traditional defense contractors and the Department of Defense--and it must be the DoD that primarily moves to improve matters. Toward that end, it will be essential to revisit and revise the business ecosystem of the Pentagon and its defense industrial base, particularly those whose equipment involves AI or AI-relevant resources such as data.

Traditional defense contractors, armed with both a deep knowledge of the acquisition process and a broad industrial base, have historically enjoyed many benefits, such as large, multi-year contracts. Though traditional defense contractors may initially face thin profit margins during production, they can, via the long tail of sustained production and aftermarket support (including subsequent engineering and maintenance services, spare part production, and other aspects), leverage the data associated with equipment to increase those margins over time. For this reason, they have contractually negotiated to define that data as their intellectual property--a situation far more consequential to this century than the last.

Going forward, this situation must change if the Pentagon is to create and sustain a substantial competitive global edge in AI. The hardware-centric acquisition process that was sufficient for the 20th century simply will not suffice in the 21st if the US wishes to sustain its competitive advantage.

While the Pentagon is well aware of the need to become AI-ready (as the NSCAI report demonstrates), it cannot succeed either alone or by working solely with traditional defense contractors. Instead, the Pentagon must create a level playing field for all contractors, including nontraditional organizations who are best positioned to deliver transformational capabilities. It must foster an open ecosystem in which innovative companies with commercially proven technologies have the opportunity to contribute to our national defense.

This will be essential not just to purchase and deploy the best available AI solutions but also to troubleshoot and optimize a wide variety of current systems acquired from other contractors. There are numerous cases where such optimization would be enriched and accelerated by AI-powered analysis.

Consider the issues involved in the F-35 fighter jet and related assets--specifically, ALIS (the F-35’s Autonomic Logistics Information System). Per the 2020 Government Accountability Office (GAO) report on ALIS, it’s fair to say the system (originally acquired via the long cycle of the hardware acquisition process, in the context of the F-35) has had a few problems. As drawn from the GAO website, we find that these problems include inaccurate data, missing data, and a suboptimal user experience, leading to a significantly slower, more costly overall maintenance process.

Rapidly addressing such issues would, however, require the Pentagon to foster a defense ecosystem in which nontraditional contractors (such as AI solution providers) can compete fairly with traditional contractors, so that AI specialists can obtain access to equipment data currently treated by contracts as the intellectual property of the traditional defense contractors.

It bears pointing out that this situation is far from unique to ALIS and the F-35. There are many other instances in which enhancing equipment or solving problems pertaining to that equipment could be bolstered and accelerated via AI, given changes to the Pentagon’s acquisition process and related contractual terms.

We should, of course, be clear that when we say “open,” we mean open competition as a means of delivering the best, most innovative technology to warfighters--not open access to sensitive data. We believe AI solution providers and other nontraditional contractors should have to prove to the Pentagon's satisfaction, prior to receiving key data, that they can and will serve as trusted partners in managing all relevant data. Complying with the rigorous terms required by the Cybersecurity Maturity Model Certification (CMMC) and the Federal Risk and Authorization Management Program (FedRAMP) are key levers to ensuring data security. If the Pentagon follows through on these recommendations, it will foster free competition and deliver on the full transformative potential of AI without compromising data security.

For these reasons, we call on the Department of Defense to revise its acquisition process, enable the fulfillment of the goals noted in NSCAI’s report by 2025, empower warfighters with the best available technology, and as a result, substantially improve the nation’s defense capabilities relative to those of foreign adversaries.

About The Author

Logan Jones is the General Manager of SGS, and has over a decade of national security and defense experience in the US and globally, in which he’s worked extensively with the DoD and with commercial defense companies.

Jones is the former vice president and founding member of the largest aerospace venture fund and innovation group, Boeing HorizonX. As senior director of business development, he also led a team at Boeing NeXt, a business division focused on the future of urban, regional and global mobility. In both roles he forged nontraditional partnerships, cultivated new business opportunities to extend Boeing’s core capabilities into existing and emerging markets, led multiple venture transactions, and launched partnerships with multinational companies and startups to create new businesses.

Additionally, Jones formerly served on the board of Wisk Aero, an urban air mobility company, and was recently chair of the urban air mobility committee at AIA.

Jones has a Bachelor of Science degree in marketing from Montana State University and a master’s degree in program management from Boston University. He also holds a master’s degree in business administration from Washington University in St. Louis.

Did you enjoy this great article?

Check out our free e-newsletters to read more great articles..