Achieving Decision Dominance: Leveraging AI in Small Wars
The information environment is expansive, complex, and rapidly evolving. During contemporary conflicts, including gray zone and hybrid warfare, perception often outweighs reality. That is why artificial intelligence is critical for navigating this complex yet fluid landscape. The Department of Defense (DoD) can enhance its effectiveness within the information environment during “small wars” to achieve decision dominance, but it needs to further incorporate artificial intelligence and its respective capabilities within its doctrine and culture.
Decision dominance ensures the commander can better understand the area of operations and deprive the enemy of the ability to make timely decisions. This method emphasizes proficiencies, not physical capability. It is not simply a matter of denying the opposing forces their decision-making ability, but rather a strategy of influencing actions by offering specific choices and limiting alternatives for the opposing force. The concept of decision dominance suggests that when an opposing force is left incapable of acting—having been stripped of all practical choices—it will cease fighting, perhaps even before major combat commences.
The DoD’s performance is in a precarious state. The last war that most closely resembled a victory was the 1991 Gulf War, and even that was convoluted, as the belligerents withdrew but remained in power. In Afghanistan in 2001, the U.S. Joint Forces won the initial stages of the war. Still, the Taliban adopted insurgent tactics and used the tribal areas of Pakistan and Afghanistan to plan their return and destabilize the environment. This led to the perception that the Afghan government was unable to protect its citizens, and the Afghan military surrendered en masse despite trillions in investment. In Iraq in 2003, the U.S. Joint and Combined Forces won the initial phase of the war and fought the later part to a stalemate.
The DoD’s experiences in both Afghanistan and Iraq highlight the challenges within the information environment. During the Iraq War, the U.S. military was inundated with vast sums of information, overwhelming operations. Decision dominance was challenging to achieve with the various competing interests in the region—and even in the Pentagon. In the information environment, a perception of victory can be as impactful as an actual outcome, while a belief in defeat can lead to tangible losses.
One of the largest challenges in both Iraq and Afghanistan was harmonizing and synchronizing practices between information-related capabilities to counter the often competitive and changing narratives of insurgents. This disconnect between official capabilities and their implementation in the field was counterproductive. The situation was further complicated by doctrinal changes (e.g., MISO), which further convoluted information-related capabilities, triggering changes in organization, practice, and doctrine.
The lack of coordination was also a key factor. The complexity resulted in setbacks because commanders were hindered from making timely decisions by insufficient synchronization and coordination of the information environment. For example, psychological operations and public affairs sometimes presented competing themes and messages, benefiting Taliban propagandists within a vastly complex information ecosystem. The Taliban routinely accused U.S. forces of needlessly causing civilian casualties. A failure to coordinate between these elements allowed the Taliban to seize the initiative and shape perceptions. This failure to shape the information environment contributed to the erosion of public trust and, ultimately, strategic setbacks.
The DoD has been investing in AI capabilities for almost over a decade. The DoD has invested in Project Maven, a sign of the DoD’s attempt to embrace artificial intelligence. However, relatively little unclassified information is known about its impact on information operations—the synchronization of information-related capabilities. The Maven prototype fuses information sources from multi-source surveillance data into a familiar battlefield analysis interface, enabling speedier and more accurate decision-making and analysis, which predictably will complement information advantage. The Maven prototype uses machine learning to sort through multitudes of intelligence, surveillance, and reconnaissance data—unmanned systems video, paper documents, computer hard drives, thumb drives, and more—collected by the department and intelligence agencies for operational use across the services.
In 2018, the DoD adopted multi-domain operations within its National Defense Strategy with a transition to near-peer adversaries. Of note, the information environment transcends all recognized war domains per the doctrine. However, challenges remain within the gray zone and hybrid conflicts, especially due to the expansive information environment. With enhancements in discriminative and generative AI, non-state actors pose both benefits and significant challenges within gray zone and hybrid conflicts, which are likely to remain below the threshold of full-blown war—especially if supported by nation-state actors with vested interests.
With the recent advent of advanced capabilities, the DoD was able to incorporate AI within its organization, culture, and operations. 2022 was a year of rapid advancement with the release of a number of public-facing models. The information environment was relatively untamed due to the failure to fully incorporate artificial intelligence into ongoing operations. With advances in artificial intelligence—thanks to Nvidia’s GPU and TPU architecture—AI is more pervasive than ever before with tools to navigate the complex information environment.
Regardless, the information environment seems to have evolved significantly since Project Maven was first introduced. With these capabilities—including off-the-shelf options—the DoD’s information operations community should enhance its campaigns and shape the information environment to achieve decision dominance and maintain information advantage to undermine the perception of the adversary in future conflicts. Information operations personnel are typically tasked with measuring the impact of information-related capabilities on a target audience. Once, that was challenging, especially before the internet and social media, making impact measurement nearly impossible.
There are a myriad of ways artificial intelligence can help tame this sprawling information environment. With internet capabilities and social media, information operations personnel can now collect observations from a computer screen and attempt to monitor the impact of information operations on a target audience. The challenge for information operations officers and personnel is gauging the sentiment of social media posts. With artificial intelligence capabilities such as discriminative AI, service members can label, classify, and perform sentiment and impact analyses using natural language processing and AI to assess their respective social media presence. With GenAI models, psychological operations can create ultra-realistic audio, visuals, and imagery. These capabilities should be gradually introduced into any information operations campaign to erode the potential for the enemy to make timely decisions.
AI has not yet been fully embraced within the DoD’s culture, organization, and doctrine. The DoD is still grappling with its effective integration to ensure that it becomes a commonly used capability. Therefore, it remains imperative for the DoD to incorporate AI within its information operations framework, operations, and doctrine, triggering what some scholars have called a revolution in military affairs. Mission-specific training must incorporate AI, and the military education system must be leveraged to ensure AI plays a prominent role in any contemporary operation—for both defensive and offensive purposes. Project Maven is a starting point, not the final product, as AI can be used for decision dominance and to maintain information advantage, especially during gray zone competition or hyper-competitive hybrid warfare.
Will these capabilities offer a silver bullet solution to achieve information advantage and prevent any more Afghanistans or Iraqs? Probably not, but they do offer the DoD the most optimal way forward to ensure that it has all the tools it needs to navigate this complex information environment—and at least align commander priorities with the operational picture on the ground.
Disclaimer: The views expressed in this report are those of the author and do not necessarily reflect the official policy or position of the Department of the Army, the Department of Defense, or the U.S. Government.
