AI in the Military: Evolution or Revolution?

Despite a centuries-old history, the technological breakthrough that our armies will face is unprecedented in the speed of its evolution

Preligens offers sensor-agnostic software that relies on artificial intelligence to detect, classify and identify military assets (aircraft, vessels, SAM/SSM, etc.)

AI has become a crucial component in various industries, transforming our lives and work. The military domain is no exception, with AI significantly impacting strategies, tactics, and operations. This article explores AI’s importance in the military, providing a comprehensive understanding of the advantages it brings on the battlefield and the necessary steps to harness its full potential in modern warfare.

The current explosion of AI is a result of the convergence of three factors: long-known algorithms (such as neural networks and deep learning), the availability of vast amounts of data, and the progress and accessibility of computing capacities. As AI continues to develop and integrate into various sectors, it is crucial to understand its significance and application in the military domain.

AI as a technological change transforming the military

The integration of technology into the military is a complex process involving economic, cultural, and social connections. Transitioning to a new mode of operation depends on internal and external pressures and available resources.

Measuring the effectiveness of technological choices is challenging due to cognitive biases. Expert judgment plays a significant role in uncertain situations, but decisiveness is crucial when facing risks.

AI in military operations has long-term implications beyond prospective concepts like “hyperwar.” Early developments, such as satellite image analysis, are essential for military advancements. It’s vital to understand how military personnel adapt to these changes and how technology fits within a complex organizational landscape, while avoiding over-optimism about the potential of disruptive technologies.

AI is a crucial issue in geostrategic competition

AI’s role in geostrategic competition should not be underestimated. The power competition that has always existed is now being influenced by the emergence of private and non-state actors outside the realm of government. The AI landscape is marked by the concentration of non-traditional defense actors, such as GAFAM, BATX, and startups working on defense AI, which together create a military-digital complex. The workings of this complex can be more visible in the U.S. due to its open political system, while they may be less apparent in China.

New technologies bring about new dependencies, with military superiority increasingly reliant on digital advancements. Countries like China are investing heavily in AI to reduce asymmetry and balance the global power dynamics. According to the Stanford University Artificial Intelligence Index Report of 2022, in 2021, the United States led the world in overall private investment in funded AI companies, approximately $52.9 billion, over three times China ($17.2 billion). The U.S. accounts for nearly 50% of private investment in AI worldwide.

This level of investment is necessary to maintain the initiative, anticipate and prevent strategic disruptions, and reap the benefits of AI as a means to achieve operational domination. The influence of AI on the military and geostrategic competition is undeniable, and countries must recognize the importance of investing in AI to remain relevant and competitive on the world stage.

Operationalization of AI: Advantages on the Battlefield

To respond to the Russian invasion in Ukraine as well as the increase in tensions in the Pacific, states are now gearing up for a form of conflict that is expected to be more massive, intensive, and sustained. The cause is clear: more personnel, more equipment, and the need for the ability to hold out for a long time. This change of direction, which also has signaled a return to the brinkmanship of nuclear deterrence, represents strong but indispensable choices.

Despite this, it would be incorrect to perceive the Ukrainian conflict as a simple reversion to the past, or an improbable return to the dialectic of blocs and direct combat. Rather, it symbolizes an ontological rupture brought about by the digitization and automation of the battlefield.

The most visible part of this transformation is the abundance of drones. There are hundreds of internet videos of these Ukrainian-piloted machines destroying Russian armored vehicles, which only represent a small portion of use cases: from the commercial versions of the Dji to the Ukrainian Bayraktar or Russian Orlan, thousands of drones are now in daily use and their application has become decisive. These autonomous piloted systems reduce the exposure of combatants and have become indispensable for observation, coordination of operations, and support through multiple targeted strikes. All this has come at a minimal cost that is all the more measured since some of these devices are simple adaptations of consumer equipment.

But the most important impact of the new technologies lies elsewhere: it is in the whole ‘C2′ (command and control) area that the effects of the digitalization of the battlefield are today the most systemic. The Ukrainians’ halt of the Russian offensive in the spring of 2022 can be largely attributed to their superior intelligence and coordination resources. The enhanced situational awareness provided by a profusion of sensors, the connectivity offered by the Starlink constellation, and the capacity to gather, analyze, and integrate a continuous flow of information have emerged as significant factors contributing to their success.

The mastery of artificial intelligence (AI) allows for a multiplication of the use of operational capabilities. The foreseeable evolution towards real autonomy of drones and swarms of drones—with the ethical questions that this raises—or the immense potential changes brought about by generative AI for the C2 function will change the way in which war is waged, but also the way in which it is planned, directed, and supervised.

Before we can fully grasp the transformation of warfare, a different change must occur. In nations boasting a modern military force, the necessity to harness cutting-edge technology and digital advancements prompts a reassessment of their ability to incorporate defense industry innovation. However, the standards-based approach and the traditionally lengthy processes characteristic of military administrations are currently ill-suited to the swift pace of technological advancements and the operational dynamics of the software economy. Thus, a profound overhaul of acquisition systems is an essential step for NATO armies.

Despite a centuries-old history, the technological breakthrough that our armies will face is unprecedented in the speed of its evolution and in the haste of the changes it implies for operational and acquisition models. The magnitude of this challenge will undoubtedly lead states to imagine new modes of public-private cooperation.

The How: Key Steps to Foster AI Development in the Military

  1. Adequate Budget Allocation: The military must address the current underinvestment in emerging technologies compared to traditional programs. This approach will enable the amplification of private sector investments’ impact and guarantee that financial commitments are fulfilled promptly and efficiently. One recommendation is to pool funds for specific purposes, such as AI, that are not tied to a particular program, allowing program managers to respond quickly to new opportunities in rapidly changing technology areas.
  2. Embrace Experimentation and Co-Innovation: Agile project management is crucial to integrate digital technologies into military operations. By engaging in real-world experimentation and co-innovation with private sector partners, the military can accelerate the development and deployment of AI solutions. Examples of this include the French PRL-TAIIA experimentation and the US Task Force 39/59/99, AFWERX, and DIU initiatives.
  3. Structuring Ecosystems for Scalable AI Development: The military needs to address data compartmentalization, standardization, and quality improvement, which are prerequisites for developing algorithms for various defense applications. Additionally, cloud infrastructures should be leveraged to support AI development. The US Department of Defense’s call for tenders from Microsoft, Google, Amazon, and Oracle for federal cloud infrastructure deployment exemplifies this approach.

Lastly, it is essential to foster close relations between the private and public sectors to catalyze the adoption of AI in the military. This can be achieved through strategic influence, collaboration between individuals from both sectors, and cross-investment in research funding to maximize the leverage effect.

By considering these potential scenarios and taking the necessary steps, military leaders can navigate the challenges of AI integration and pave the way for a future where AI serves as a powerful and responsible tool in the hands of the armed forces.

Conclusion

In conclusion, integrating AI into the military offers opportunities and challenges. Realizing its potential advantages in operations, strategy, and geostrategic competition requires overcoming skepticism, institutional barriers, and budget constraints. Understanding the complex socio-technical landscape enables military leaders to navigate these challenges for successful AI integration.

By fostering collaboration between private and public sectors and addressing challenges, military leaders can ensure AI becomes a powerful, responsible tool, enhancing capabilities and maintaining initiative, strategic anticipation, and operational domination in a competitive geostrategic landscape.

Main image: Preligens offers sensor-agnostic software that relies on artificial intelligence to detect, classify and identify military assets (aircraft, vessels, SAM/SSM, etc.)

Posted in: Contributed   Tagged in: Show Daily - May 22

An Introduction to Radar Imaging

What it is and how it works