The U.S. Department of Defense recently unveiled an ambitious initiative aptly named Replicator, aimed at rapidly scaling capabilities in the face of strategic competition with the People’s Republic of China. Replicator’s first task will be to quickly scale and field thousands of attritable autonomous systems within the next 18 to 24 months, leveraging AI, robotics, and commercial technology. This initiative is the latest in a series of institutional pushes the Department of Defense is making to transition advances in emerging technologies into realized and ready-to-use capabilities. The initiative’s intent is to keep pace with China’s efforts to “intelligentize” its military by leveraging an array of cutting-edge technologies to pursue its foreign policy goals in the Indo-Pacific.
Replicator’s goal is to integrate emerging technologies — and particularly those originating in the private sector — into the military’s operational framework. However, there is growing concern that the Department of Defense’s recent initiatives and existing processes may not be sufficient to meet the immediate challenges, that the planned degree of change isn’t sufficient, and that the department risks falling down a path of risky incrementalism. The planned changes, such as the launch of a generative AI task force or efforts to conduct extensive AI training and education of the Department of Defense workforce, are long-term investments that won’t yield immediate results. This leaves the United States potentially vulnerable in the short term as China continues to rapidly build up to blunt current U.S. operational advantages.
The question remains whether the pace and scale of Replicator’s innovations can meet the demands of an increasingly complex and competitive geopolitical landscape. If Replicator lives up to its hype, it could create a streamlined pathway for integrating emerging technologies into the military — another cut at the bottleneck analysts and policymakers alikehave been lamenting has hindered the U.S. military’s ability to maintain its competitive edge. So, while the promise of Replicator is immense, its success hinges on overcoming a myriad of challenges, from production scalability to bureaucratic inertia, that have hindered previous similar innovation adoption efforts.
From Exquisite to Attritable
For decades, the U.S. military has grappled with the challenge of balancing the cost and effectiveness calculus between the “many and simple” and the “few and complex.” Since the Vietnam War, the U.S. military has leaned toward the latter, prioritizing expensive, exquisite systems, best exemplified by the pursuit of more precise weapons — built in smaller numbers — to maximize the utility of extremely capable platforms. This approach stemmed from a basic reality: The United States couldn’t match the Soviet Union’s production of conventional military equipment. Therefore, the United States would invest in technology to make its forces more lethal, such as the previous effort to field an array of increasingly “smart” weapons systems — dubbed Assault Breaker — in the mid to late Cold War. However, ballooning costs for weapons and platforms, recent breakthroughs in emerging technologies, and the re-emergence of great power conflict make it increasingly difficult for the U.S. military to rely solely on small numbers of advanced capabilities.
Attritable systems can offer a viable solution to this cost curve challenge because cost-effective platforms, networks, sensors, and munitions can be mass-produced. These characteristics — particularly their low cost — make them particularly well suited for attrition warfare and would allow for greater operational flexibility. By branching out from a sole dependency on a limited number of expensive, often crewed, assets to include and make space for complementary, more cost-effective, uncrewed systems that can be produced, deployed, and redeployed with less maintenance, the United States can innovate through reintroducing mass into its military planning.
Attritable systems’ strategic advantages are greater than the sum of the individual systems. These platforms and systems can “disaggregate expensive … systems,” which allows for the users to disperse combat power, forcing adversaries to use more munitions to defend. The use of large numbers of cheaper systems also increases the chances for “graceful degradation,” meaning the loss of a single system is significantly less detrimental to overall combat power. At a systematic level, therefore, the use of these sorts of systems shifts the burden from an individual system’s survivability to the overall resiliency of the swarm or network.
Technological Readiness Meets Geopolitical Imperatives
Replicator, and the focus on all-domain attritable autonomous systems, arrives at the crossroads of three key recent developments: 1) growing technological readiness, 2) clear proof of concept, and 3) geopolitical necessity.
First, AI, robotics, and autonomy have matured to a point where they can be safely and reliably integrated into military operations. Many applications are not only sufficiently mature but also increasingly accessible — largely because the primary drivers of much of the underlying technology are not military- or government-affiliated research institutions, but rather academia and the private sector. These technologies are not just cost-effective. Instead, they are often general-purpose, if not dual-use, with many of them nearly ready to use straight off the shelf. Such accessibility and affordability offer militaries more room to experiment, fostering tactical innovation.
From the Navy’s Task Force 59, which, as Schuyler Moore discussed on the War on the Rocks podcast, operates essentially a small flotilla of autonomous systems in the Arabian Gulf, to the Air Force Research Laboratory’s Golden Horde swarming program, established experiments have showcased the readiness, applicability, and opportunities of these systems. Task Force 59, in particular, illustrates some of the kinds of capabilities (albeit slightly more expensive) that Replicator could accelerate the delivery of (if done well). However, Task Force 59, despite many successes, has faced institutional delays in scaling up — with private companies being underutilized and contracted to deliver only a fraction of their delivery capacity.
Replicator’s focus on attritable mass, therefore, marks the graduation of these emerging capabilities from experiments to force elements, arguing that not only is the technology ready, but also the Department of Defense is ready to procure them quickly and effectively integrate them with how the military fights. Moreover, there are accompanying policies already in place, including Directive 3000.09, the Responsible AI Strategy and Implementation Pathway, and the Department of Defense’s Ethical AI Principles. The Department of Defense has been a leader in developing policies to safely and responsibly develop AI and autonomous systems.
However, having mature technologies is not enough. There often exists a “trust gap” between the true and perceived abilities and limitations of a given technology. It is important to keep in mind the difference between technological change and military innovations that necessarily require organizational change and shifts in how militaries fight. It was only until recently that proof of concept was so obviously and dramatically established.
Often, however, technology reaches an inflection point, or a critical demonstration to push it over the finish line from a theoretical capability to an actual deployed system to a system scaled across a military. Ukraine’s successful and wide-scale demonstration of the benefits and opportunities afforded by leveraging mass on the battlefield has finally catalyzed the absorption of “lessons learned” from the conflict. Through tracking and queuing targets through a specialized app — dubbed the “Uber for artillery” system — to Ukraine’s development of an “army of drones,” Kyiv has demonstrated some of the potential for scaling cost-effective, off-the-shelf attritable systems, including, most recently, the Ikea-style, flat-packed, build-it-yourself cardboard drones that cost no more than $3,500 each. This is especially true when these systems are well integrated and interoperable with the rest of the military. The ongoing conflict in Ukraine serves as a real-world testbed, demonstrating the effectiveness of using attritable systems such as drones as expendable resources.
The Ukrainian military’s use of drones in large numbers — and, perhaps more significantly, its ability to withstand staggering losses of these systems with seemingly constrained impact on its ability to fight — has shown that quantity, when strategically employed, can indeed have a quality of its own. Russia’s deployment of waves of relatively unsophisticated, easily destroyed Shahed loitering munitions, for example, demonstrates the effective cost asymmetry; even if they are easy to shoot down, the payoff might not justify the effort — or the cost.
Finally, the resurgence of great power conflict — especially in the face of China’s clear advantage in mass, with “more ships, more missiles, more forces” — has provided the impetus and urgency needed to continue shaking up the budgetary, organizational, and bureaucratic status quo. Replicator aims to accelerate the deployment of advanced, attritable autonomous systems, shifting the balance of power by making some emerging capabilities and operational concepts thought to be a decade away a reality much sooner.
A High-Risk, High-Reward Gamble
Deputy Secretary of Defense Kathleen Hicks described the initiative as a “big bet” — indeed, the promise of Replicator is immense; it proposes to do something the Department of Defense hasn’t done before. However, the department has struggled with these “big bets” in the past.
One of the most formidable challenges lies in scaling up production. The U.S. industrial base will need to adapt quickly to meet the demand for mass-produced, cost-effective military assets. The Department of Defense should act decisively to get the most impactful technologies into the hands of operators and warfighters as soon as possible. This is not just about speeding up production of systems, but also about accelerating the entire lifecycle — from experimentation and testing to evaluation — to ensure these capabilities can be deployed effectively.
Bureaucratic inertia also poses a significant risk. The Department of Defense has a history of promising innovation but falling short due to a labyrinth of institutional barriers and bureaucratic red tape. Budgeting processes can be so protracted that, by the time the funds are proposed by the Department of Defense and appropriated by Congress, the technology is outdated (an acute risk given how quickly the field of AI is evolving). The military services sometimes resist new technologies whose integration requires new concepts of operation that challenge existing bureaucratic hierarchies. Innovation offices that experiment with and prototype new technologies are often disconnected from procurement decision-making in ways that extend technology transitions or even prevent it from occurring.
If Replicator works, the initiative will compress the timelines for procurement by focusing on a known operational challenge — the Chinese military — and capabilities that are almost ready or ready to scale. Hicks has already indicatedthat the initiative would not need new money or bureaucracy, since it will leverage existing stakeholders, mechanisms, resources, and levers to implement it — suggesting that enough is in place that the “attention of leadership” (via the Deputy’s Innovation Steering Group) will provide the impetus for Replicator to succeed in scaling thousands of attritable, autonomous systems.
The Deputy’s Innovation Steering Group should quickly identify and prioritize those capabilities that are already ready to scale, and also capabilities that, with accelerated experimentation, testing, and evaluation, could be ready to scale. The focus should remain aligned with the National Defense Strategy if this initiative is to be successful in achieving its broader aims. Moreover, the innovation steering group has the potential to become a mechanism for the further integration of the Defense Innovation Unit throughout the capability development process, to hopefully achieve the best of both worlds: capabilities informed by long-term defense research, development, testing, and evaluation and infused by rapidly advancing commercial technology.
Moreover, with a relatively straightforward set of publicly announced metrics (thousands of systems) and an observable, short timeframe (18–24 months), the initiative must demonstrate quick successes to maintain momentum. The urgency to field these technologies is high, and so are the stakes. Moving too slowly in the early stages could not only derail the initiative but also erode trust in the feasibility of rapidly integrating advanced technologies, risking backsliding on these types of military innovations.
Replicator, therefore, needs an early win to demonstrate that rapid scaling is possible to the bureaucracy and industry. The Department of Defense therefore should select a specific system viewed as ready to scale, fund it, and announce it to send a signal to both external industry players and internal stakeholders that government is committed to this initiative.
Conclusion
In response to the formidable threat from China in the short term, the Department of Defense has wisely zeroed in on accelerating its ability to scale and leverage existing systems via the Replicator initiative, starting first with thousands of attritable, autonomous systems within the next two years. This initiative is a crucial step in turning technological innovation into military innovation, harnessing advances in AI and autonomous systems to bring capabilities thought to be accessible in the 2030s and beyond into the U.S. military much faster. Arguably more important than this first few thousand drones, if all goes well, the Replicator methodology could then itself be replicated and applied to future efforts to build the necessary bridges for military and commercial-sector technologies to cross that “valley of death” (the potential gap technologies face when transitioning into a program of record). While challenges are inevitable and a healthy skepticism exists of the ability of the Department of Defense to execute the initiative within the existing budgets and bureaucratic frameworks it already has, early commitment and demonstrated success are essential for both internal and external validation. Ultimately, Replicator holds the promise of not just countering China’s advancements in the short term, but also of fostering a more agile and innovative American military in the long run.
Lauren Kahn is a senior research analyst at Georgetown’s Center for Security and Emerging Technology, focused on national security applications of artificial intelligence. Twitter: @Lauren_A_Kahn
Image: Pvt. James Newsome, 11th Armored Cavalry Regiment
Comments are closed.