David Spirk served as the Pentagon’s Chief Data Officer from June 2020 to March 2022. In the following analysis, he lays out four ways the department needs to rethink how it approaches warfare in the modern era, with an eye on treating data and IT as vital warfighting tools.
The dismaying events on the European Plain make it clear the Department of Defense (DoD) has hard decisions ahead that are necessary to adapt to the rapidly changing digital character of war. While much of the attention on those events has rightly focused on Russian conventional forces’ performance via territorial gains and losses, the DoD must not become distracted from the pacing threat.
Future victory will go to the side that most quickly applies overwhelming force while dispensing truth to maintain an honest global representation of events. The DoD must move faster with a common understanding there are now totalitarian adversaries building massive digital arsenals across their diplomatic, informational, military, economic, financial, intelligence, and law enforcement institutions.
The new state of conflict is a yet-to-be-fully-conceived blend of strategic weapons, tactical operations, and dominance in information warfare. In this future, data is the exponential asset, appreciating in strategic value as new forms are collected and connected — and a new warfighting paradigm is needed to be competitive in this strategic landscape. The DoD must courageously mobilize the human capital, resources, commercial partnerships, and creativity of the national security community to meet this challenge.
There are four areas the DoD must emphasize to accelerate data-driven warfighting:
1) Focus on talent, not tools: Building new architectures, processes, and tools take time, but true success comes when operators and analysts become fluent in data-driven technology. Inserting the best available data-driven technology on a rolling, regular basis — call it “relentless incrementalism” — will enable operators and analysts to find creative ways to join data and models to create new operating concepts that will drive necessary doctrinal change. As operators become more proficient, there will be an increasingly vocal demand for broad improvement across the data ecosystem (e.g., people, technology, and culture) that should be encouraged.
Furthermore, the DoD must access the best talent, which requires non-traditional partners from across the commercial innovation base. This includes a new generation of leaders working across the civil service, contract support services, and industry/academic partnerships. The DoD must enable every person serving in the DoD to develop a degree of data fluency to integrate with their domain expertise. Accordingly, leaders across echelons must model the importance of data-driven decisions; DoD executive education should include a two-day data boot camp, and the department should institute a “Data and AI Day” to promote data fluency and data-driven applications across the workforce. Ideally, the military departments would multiply the positive benefit through the addition of digital and data acumen to training, promotion criteria, and institute regular guided learning activities at echelon (e.g. A modernized “Command Run”).
2) Procure a flexible data architecture. The DoD must rationalize it’s data architecture with modest centralization and a lot of consistency (e.g., abiding by the DoD’s Data Decrees). The department must obstinately guard against legacy concepts of a single-data standard or a common data model, lest the defense industry evolves to take advantage, once more putting DoD behind the commercial data curve; instead, the department must embrace the flexibility provided by a commercial minded open-data standard architecture. The goal is bold but simple: true commercial cloud computing capability has to be available for every warfighter or platform at the edge operating in a disconnected and disadvantaged communication environment. It is at this edge that our warfighters will develop data-driven concepts and technologies for maximum speed and proficiency that will eventually drive doctrinal change
To this end, it is more clear today than yesterday that information technology (IT) is a warfighting capability, and thus Chief Information Officers (CIOs) and the acquisition community need to treat it just like any other urgent battlefield need: with rapid product procurement and employment. DoD CIOs need to be resourced and enabled to move beyond keeping old systems working, quarterly software updates, staffing authority to operate packages, and running help desks. The department’s IT professionals are warfighters and the DoD must empower them to ensure commercial cloud services are continuously upgraded to best-of-breed hardware, while avoiding oppressive cyber security requirements that make agile software and model development with commercial industry and academic partners impractical. And to truly fight in a coalition environment, DoD needs seamless bi-directional data flows across classifications with allies and partners at the speed of commercial compute and modern wartime decision-making, not merely an incrementally better version of the same physically secure systems that we have today.
All this is achievable with a movement to zero trust that prioritizes investment and rapid integration of cryptographic isolation and post-quantum encryption technology over improving legacy concepts of physical separation and maintenance of multiple networks.
3) Make agile investments: Much as financial markets developed new asset classes (e.g., venture capital) to drive returns in the face of rapid and broad technological change, the DoD must revisit its resource allocation mechanisms. Senior leaders should go to Congress and seek authorities to establish monetary incentives to spur a partnered cadre of DoD, commercial, and academic technology leaders to outperform each other in investing in and deploying technology that allow the reimagining of entire warfighting concepts rather than the incremental improvement of current capabilities. The DoD must be able to make big and little bets against the future commercial data technologies, processes, and concepts that will radically change the battlefield. Successful venture capitalists with a proven track record of investment in data-driven tech should be called to public service and the DoD should empower them with the resources to make a difference. The only measure of success is the rapid (months, not years) transition of cutting edge capability to live missions where warfighters can sense, make sense, and act with greater speed and precision.
This type of flexible investment can only be made with Congressional support for an adaptive budgeting framework structured around the characteristics of an item, not the lifecycle phase [e.g. evolving items (software), enduring items (heavy machinery), and expendable items (services, munitions)]. The DoD should ask for an updated budget process that trades annual programming cycles for more frequent updates on flexible technology acceleration portfolios. These capability portfolios budgets would coordinate technology investments that cut across multiple weapons systems and the integration of cross-service resources around key operational challenges such as Joint All Domain Command and Control (JADC2).
4) Institute a continuous campaign of learning: The DoD must expand the ability for experts to persistently integrate cutting edge data-driven technologies in our deployed and exercising operational elements. The DoD now has several examples of the integration of data-driven technologies to radically transform the speed and precision with which warfighters can successfully accomplish their missions. By investing in the pathfinding efforts such as Project Maven—the DoD’s pathfinding AI effort—we are gaining a vastly improved understanding of the benefits of commercially available technology that domain experts are rapidly integrating and adapting into their live formation. Success of this model is apparent in the Special Operations enterprise, through US Northern Command’s Global Information Dominance Experiment, at the Army’s 18th Airborne Corps’ Joint Exercise Scarlet Dragon, the Navy’s creation of Task Force 59, and the Air Force’s integration of algorithmic model’s from its AI Accelerator.
This direct industry partnering with operators facilitates the real-time identification of opportunities for further integration and adaption for ever improving capabilities to support mission success. Arguably more importantly, it allows the Office of Secretary of Defense and Joint Staff a chance to identify and remove legacy policy and processes obstacles with surgical precision and bring the DoD ever closer to industry’s ability to adapt to changes to an ever changing worldwide operating environment of opportunities and threats. The Deputy Secretary of Defense’s launch of the Artificial Intelligence and Data Accelerator initiative and associated deployment of Operational Data Teams to the 11-Combatant Commands and Joint Staff is a powerful example of the DoD’s intent to ingrain this campaign of learning across tactical, operational, and strategic levels.
In summary, the DoD, the United States interagency, and our allies and partners are currently challenged by nation states, criminal networks, and extremists, all whom have an ability to leverage commercially available data driven technologies for strategic effect. To stay ahead, the DoD must bend the curve of data driven transformation faster by driving the data-driven technology into its real-world formations from the most tactical edge all the way to the halls of the Pentagon. The rapid evolution of data-driven capabilities from industry are outpacing the Pentagon’s ability to adopt and adapt them for advantage to make better and faster decisions than our adversaries. A foundational investment is needed across the DoD’s data ecosystem (e.g., people, technology, and culture) to enable the DoD to preserve its military edge into the future.
The good news is that the department has recently taken a bold move in the right direction. The establishment of the Chief Digital and Artificial Intelligence Office (OCDAO) will help the entire DoD to prioritize, resource, and scale leading data-driven technology efforts with a policy arm to reduce challenges to acceleration. This approach will enable a new generation of digital and data leaders, with both technical and warfighting domain expertise, to expand and align initiatives to the DoD’s strategic objectives. It will enable the DoD’s current data architecture to evolve to meet future needs through a close partnership with CIOs spanning the DoD’s Components. And it will encourage the emergence of a flexible investment instrument will facilitate the integrations of increasingly available commercial data-driven technology and concepts in its real-world formations.
Only disruptive change at scale in our real-world formation will make competitors unbalanced and outpaced by American ingenuity and imagination. Big decisions and associated resourcing investments and divestments are required. It is an aggressive approach, but there are no easy alternatives to meet the challenge.
Dave Spirk is the former-Chief Data Officer (CDO) of the Department of Defense. Previously, he served as the first CDO and Director of Artificial Intelligence at the United States Special Operations Command. For over two-decades, he served/deployed worldwide as a Federal Civilian in the DoD/Intelligence Community and as a United States Marine. He is currently an advisor to CalypsoAI.