Artificial intelligence (AI) is changing the future of battle. The warfighter’s tools will include capabilities that enable quick, agile decisions and maneuvers with increased accuracy. Soldiers, Sailors, Marines, Guardians and Airmen will be equipped with communications networks and modeling and simulation platforms that rapidly aggregate and analyze data, allowing for real-time decisions. Autonomous systems will operate independent of external control for progressively more extended distances and times.

The United States Department of Defense (DOD) is working with Allies and Partners and industry to develop and scale AI technologies to counter pacing threats — including China and Russia — and ensure global security. As AI technology rapidly evolves, militaries race to deploy new capabilities while grappling with the potential risks these tools bring to the battlefield. While talk of using AI in weapons can conjure movielike scenarios where mutinous robotic soldiers turn against humankind, defense experts say the greater risk lies in not using the technologies enabled by AI to strengthen defense and deterrence.

“As we focused on integrating AI into our operations responsibly and at speed, our main reason for doing so has been straightforward: Because it improves our decision advantage,” then-U.S. Deputy Defense Secretary Kathleen Hicks said while unveiling the DOD’s updated AI strategy in November 2023. “From the standpoint of deterring and defending against aggression, AI-enabled systems can help accelerate the speed of commanders’ decisions and improve the quality and accuracy of those decisions, which can be decisive in deterring a fight and winning in a fight.”

The Mines Eye surveillance drone system uses magnets, optical sensors and artificial intelligence to detect land mines in the Kharkiv region of Ukraine in October 2024. Getty Images

AI technology is the simulation of human intelligence processed by machines, typically computer systems, by collecting and classifying information that lets the machine perform specific tasks, such as writing a report or transcribing data. Machine learning is a branch of AI that enables computers to learn without external programming. These systems use data to self-train and find patterns or make predictions, growing more accurate over time as the system collects and integrates more data. Examples include predictive text on a mobile phone or an online shopping site that makes suggestions based on a user’s past purchases.

In addition to quicker, more accurate decisions in the battlefield, AI tools can assist the military in other ways. Training platforms, autonomous vehicles, logistics tools, and intelligence and cyber defense capabilities incorporate AI technology. Satellite imagery can use predictive AI tools to help pinpoint locations where security threats may occur, including deployment of nuclear weapons and other weapons of mass destruction.

“Analysis is really great, but it’s mainly retroactive, a forensic capability of looking back in time,” William Marshall, chief executive officer of Planet Labs, a global satellite imagery company, told Defense One magazine in June 2024. Planet’s satellite imagery helped U.S. intelligence officials identify Russia’s buildup of troops, aircraft and weapons in Ukraine’s Crimean Peninsula preceeding Russia’s 2022 invasion of Ukraine, Defense One reported. “In principle, generative AI models … can leverage satellite data to predict what is likely to happen: ‘You’re likely to have a drought here that might lead to civil unrest,’ ” Marshall said.

AI is being used to share data across the joint force and with Allies and Partners. The DOD launched Project Maven in 2017 to develop AI capabilities by leveraging technology to automatically identify targets on the battlefield. The department’s 2018 National Defense Strategy identified AI as one of the key technologies that will ensure the U.S. “will be able to fight and win the wars of the future.” Australia, the United Kingdom and the U.S. have incorporated AI algorithms on multiple systems to enhance data sharing and processing among their antisubmarine, reconnaissance, precision targeting and other capabilities. U.S. Strategic Command (USSTRATCOM) will incorporate AI into its nuclear command, control and communications enterprise.

An attendee wearing a virtual reality headset readies a weapon for a war simulation at the Defense Innovation Forum in Paris. Getty Images

“Advanced AI and robust data analytics capabilities provide decision advantage and improve our deterrence posture,” said USSTRATCOM Commander Gen. Anthony Cotton. “IT and AI superiority allows for more effective integration of conventional and nuclear capabilities, strengthening deterrence.”

It’s important to note that AI will always have human oversight and merely assist with data collection and analysis to gather and integrate information more quickly, develop faster solutions and provide more decision space to USSTRATCOM leadership.

“Everything we do has a human in the loop,” Vice Adm. Brad Cooper, deputy commander of U.S. Central Command, told an audience of military leaders, government officials, industry representatives and scholars at a March 2024 conference in Tampa, Florida, presented by the Global and National Security Institute at the University of South Florida. “At the end of the day, decisions are made by humans. It’s the decision-making process that is more vibrantly enabled through AI. We’re able to move at speeds that were previously unimaginable.”

Washington is working to ensure the U.S. stays ahead of competitors in developing AI capabilities. In 2017, China released a strategy detailing its plan to take the global lead in AI by 2030. Less than two months later, Russian President Vladimir Putin announced Russia’s intent to pursue AI technologies.

“I think the global competition has got to be first and foremost in our minds at all times,” U.S. Sen. John Hickenlooper said during a Center for Strategic and International Studies (CSIS) panel in November 2024. He noted that China is investing a “staggering” number of resources into building the country’s AI capabilities. “But we have a system in this country of entrepreneurship and innovation tied together that, with a collaborative history between that … our institutions of higher learning, our private sector, our military, they’re all … accelerating the advances we make.”

In October 2024, the White House issued a national security memorandum on advancing U.S. leadership in AI, laying out policy objectives for accelerating adoption of AI systems.

“We have to get this right, because there is probably no other technology that will be more critical to our national security in the years ahead,” then-U.S. National Security Advisor Jake Sullivan said during remarks at National Defense University in October 2024. “The stakes are high. If we don’t act more intentionally to seize our advantages, if we don’t deploy AI more quickly and more comprehensively to strengthen our national security, we risk squandering our hard-earned lead. Even if we have the best AI models but our competitors are faster to deploy, we could see them seize the advantage in using AI capabilities against our people, our forces, and our Partners and Allies. We could have the best team but lose because we didn’t put it on the field.”

We have to get this right, because there is probably no other technology that will be more critical to our national security in the years ahead.”

Jake Sullivan, then U.S. national security advisor

The decision advantage Hicks emphasized in the 2023 AI adoption strategy focuses on five outcomes: (1) battlespace awareness and understanding; (2) adaptive force planning and application; (3) fast, precise and resilient kill chains; (4) resilient sustainment support; and (5) efficient enterprise business operations. The department aims to achieve each of these outcomes with an “agile approach” to adoption, ensuring a tight feedback loop between technology developers and users through a continuous cycle of iteration, innovation and improvement to accelerate deployment speed measured in hours or days, not months or years. “Creating effective, iterative feedback loops among developers, users, subject matter experts, and test and evaluation experts will ensure capabilities are more stable, secure, ethical and trustworthy,” the report reads. “An agile approach to adoption emphasizes speed of delivery and continuous improvement, prioritizing outcomes over processes.”

The energy challenge

Ensuring timely deployment and dependable availability of AI tools for the defense and other sectors requires a massive buildup of energy infrastructure. Data centers needed to store massive amounts of digital information take up large swaths of land and increase energy use by orders of magnitude. In addition, servers that operate AI software require a lot of water for cooling. There’s a clear environmental impact: Google and Microsoft reported in 2024 that their companies saw double-digit increases in emissions during the past several years, driven by new data center energy use to support AI workloads.

Using an AI tool for a basic text query can require 10 times more energy than a Google search, according to industry experts.

 “One query to ChatGPT uses approximately as much electricity as could light one light bulb for about 20 minutes,” Jesse Dodge, a senior research analyst at the Allen Institute for AI in Seattle, told National Public Radio. “So, you can imagine with millions of people using something like that every day, that adds up to a really large amount of electricity.”

Neil Chatterjee, former chairman of the Federal Energy Regulatory Commission, said energy demands for AI tools will need to be met with a mix of traditional and renewable energy sources to keep pace.

“We’ve got to get our demand projections accurate and figure out how much power we will need,” Chatterjee said. “I don’t think we have fully started to wrap our brains around what it will take to meet this coming surge in demand while maintaining not just affordability of power, but reliability as well. We’ve got to make sure that we’ve got a sufficient amount of power not just to support data centers for AI, to win the AI race, but to also make sure that residential consumers in the middle of a really hot summer or a really cold winter don’t have power curtailed.”

Ukraine has used the Turkish Bayraktar TB2, a medium-altitude, long-range combat drone, in its war against Russia.

Demand for AI capabilities, as well as the energy infrastructure to enable them, is best met through partnerships between government, industry and Allies, said Chris Lahane, vice president of global policy for OpenAI, developer of ChatGPT. He noted that the U.S. has succeeded in developing massive infrastructure projects in the past, including the nationwide interstate system and early development of the internet.

“Today the U.S. is winning but that lead is not guaranteed,” he said during the CSIS panel. Keeping that lead will require progress in multiple areas, including construction permits, investment incentives and reinvigorating nuclear power. “This is really a time where we need to start to think big given what’s at stake. Again, there’s two nations in the world that can build this stuff at scale and it’s the U.S. and China.”

Establishing guardrails

Some leaders worry that emerging AI capabilities will increase risks to safety and stability. In a July 2023 report, United Nations University called for an AI governance framework that will adapt as needed along with technology to avoid risks including bias, privacy and security. AI could be used to target groups or individuals based on faulty data gathered by commercial tools and accessed by government.

“Data is the fundamental issue here. It causes a huge amount of concerns and vulnerabilities, and it’s not being accounted for when we talk about things like traceability broadly,” Meredith Whittaker, president of the Signal Foundation, a nonprofit organization that advocates for private, secure communication, told Defense One magazine. Whittaker is one of three authors of an October 2024 paper published by the AI Now Institute at New York University calling attention to potential risks of AI. Because AI relies on patterns gleaned from public and personal data to perform functions, innocent civilians could erroneously be placed on a target list, according to the paper. The authors urge developers to insulate military AI systems and personal data from commercial models. She called on government to broaden privacy laws and strengthen laws to prevent bad actors from gaining access to commercial datasets that include civilians’ personal information.

Participants gather at the Responsible AI in the Military Domain summit in Seoul, South Korea, in September 2024. Getty Images

In its AI security memorandum, the U.S. government directed DOD officials and the intelligence community to examine how existing policies and procedures affecting privacy and civil liberties can be revised to “enable the effective and responsible use of AI.” The memorandum also calls for other federal agencies to take active steps to uphold human rights, civil rights, civil liberties, privacy and safety.

“We will succeed by leading with our strengths: our democratic values, our diverse and open society, our culture of ingenuity, our second-to-none innovation base, and our globe-spanning network of Allies and Partners,” Hicks said. “Together we will harness data, analytics and AI for the defense,
security and prosperity of the American people and the world.”  

Share.
Leave A Reply