How Agentic AI Can Transform Military Decision-Making

The Department of Defense should actively begin experimenting with Agentic AI tools in the Joint Operations Planning Process (JOPP)—an AI that can autonomously complete a series of tasks to achieve specified complex objectives. [2] There are two reasons for this. First, compared to humans, Agentic AI can more quickly and comprehensively synthesize a wide range of traditional and non-traditional planning factors, thus helping to formulate more comprehensive and objective Courses of Action (COA). Second, once a COA is selected, Agentic AI can also assist in rapidly issuing downstream instructions and commands, streamlining communications, and saving hundreds of work hours in each planning cycle.

Agentic AI is a capability that can rapidly respond to changing battlefield conditions and assist in independently solving large-scale complex problems. This is different from the currently popular large language models, which rely on single prompts to perform simple, specific tasks. Creating multiple dilemmas for equally matched adversaries requires the continuous integration of all force tools and capabilities across all domains, including the electromagnetic spectrum and information environment. [3] In the Fourth Industrial Revolution, Agentic AI is a method of deploying various autonomous-based technologies that work together to perceive their environment and autonomously determine action plans to achieve established goals. [4] Combining this technology with human planners can produce an accelerated multidisciplinary thinking machine.

Imagine a planning team with a multifaceted “agent” that can understand geopolitical trends, global dynamics, and national policies related to conflicts. It can also interpret the limitations and constraints of military forces across all operational domains by investigating multiple datasets. This “thinking spear” can also minimize the impact of groupthink, self-interest chasing, and counterproductive biases, creating new opportunities and pathways for decision-makers. Deputy Secretary of Defense Kathleen Hicks confirmed this point during the unveiling of the Pentagon’s 2023 Data, Analysis, and AI Strategy, stating, “From the perspective of deterring and defending against aggression, AI systems can help accelerate commanders’ decision-making speed and improve the quality and accuracy of decisions.” [5] We propose that Agentic AI is a new frontier “AI enabler” that the Department of Defense should accelerate its adoption of to achieve these goals.

Alternatively, envision the United States—slow to adapt and constrained by traditional planning processes—competing at the strategic, operational, and tactical levels with an adversary equipped with this “thinking spear.” No amount of high-tech equipment in our soldiers’ hands can overcome an adversary that has better and faster information flow, allowing them to win with intelligence. Competing with an adversary that has this intelligence and decision-making space is a consideration worth seriously pondering the use of Agentic AI in parallel planning structures.

The Russia-Ukraine war has shown us the value of AI in modern warfare and its impact on military operations and tactics. Earlier this year, TIME magazine reported that Palantir Technologies’ AI software was responsible for much of Ukraine’s targeting. [6] Additionally, Palantir assigned a software engineer to each battalion, demonstrating that this experiment accelerated “the most significant fundamental change in the nature of warfare ever,” as stated by former Chairman of the Joint Chiefs of Staff General Mark Milley. [7] In fact, Defense One reported that the Pentagon has also been integrating “AI and machine learning into its intelligence, surveillance, and reconnaissance operations, helping Ukrainian forces thwart some Russian attacks.” [8] These emerging experiments with AI on the battlefield indicate that our military also urgently needs to gain an edge in the decision-making process.

Agentic AI in the Joint Operations Planning Process can provide an information advantage at a relevant speed. Next, we propose several ways Agentic AI can serve as an effective means to achieve objectives:

• Agentic AI with exceptional multi-domain awareness can suggest troop posture recommendations to planners and create multiple dilemmas within the Multi-Domain Operations (MDO) construct, as it can continuously manage information on the actions of joint and coalition forces as well as adversaries.

• Agentic AI can assist in prioritizing the Joint Integrated Prioritized Target List (JIPTL) based on real-time battlefield conditions, including the adversary’s capabilities, approaches, risks, and opportunities.

• Agentic AI can track and identify potential logistical shortages (e.g., fuel, supplies, ammunition) to ensure consistent support for dispersed forces across a vast theater of operations.

• Agentic AI can make “understanding the enemy” central to COA development. The red team is an element that planners may quickly overlook, as the pressures of conflict naturally lead people back to familiar ways of fighting without considering the enemy’s input.

• Agentic AI can instantly synchronize guidance and intent across the entire battlefield, reducing the likelihood of collateral damage while enhancing tactical flexibility and lethality.

• Finally, and most fundamentally, planners can leverage AI to create and disseminate all downstream orders generated during periodic planning processes, saving hundreds of hours of tedious, repetitive administrative input in each cycle and allowing more combat personnel to return to action.

We acknowledge that there is still much to learn about the risks and resilience of Agentic AI in a competitive communications environment. Theoretical discussions about ethics, security, and best practices should continue. Nevertheless, countries like China are competitive in the AI race and are clearly eager to gain a technological advantage. Future wars will almost certainly be won first in the information domain.

Military leaders should accelerate the experimentation and adoption of Agentic AI tools, integrating them into the Joint Operations Planning Process. Importantly, they should do this with an iterative mindset, striving to mitigate risks as they arise (machine learning will help in this regard) rather than waiting for a perfect product to implement. When on the brink of a technological revolution, we must accept the risks that come with taking a significant leap forward. Because, undoubtedly, the risks to national security are greater if we do not become the first major power to harness this formidable force.

This article is sourced from: Blue Science Game Theory