AI is fast becoming a critical aspect of modern warfare, but the ROI on AI investments continues to ... [+] be questioned, even in the defense sector. (Photo by Cristina Young/U.S. Navy via Getty Images)Getty ImagesIn October of this year, outgoing U.S. President Joe Biden issued the first-ever National Security Memorandum on Artificial Intelligence, which ordered the U.S. military and intelligence agencies to more quickly adopt and deploy artificial intelligence. It was a move that further showed what many global leaders already recognize: AI is now a critical aspect of modern warfare.But as defense departments worldwide rush to integrate AI into their operations, questions continue to linger on how to get the best value from their AI investments and deployments. Its therefore not surprising that the executive order wasnt just for the United States to lead the way in seizing the promise of AI, but to also manage the risks of AI.The press release from the White House also noted that the memorandum was designed to galvanize federal government adoption of AI to advance the national security mission, including by ensuring that such adoption reflects democratic values and protects human rights, civil rights, civil liberties and privacy.For Meir Friedland, CEO of RF spectrum intelligence company Sensorz and an expert in AI for military tactics, AIs potential in defense is massive, but achieving it requires a calculated approach. AI stands at the forefront of modern warfare, bridging physical and digital domains. But without clear strategies and rigorous safeguards, its potential can easily go unrealized or worse, backfire, he said.Thats the realization thats dawning on defense departments across the world. What lies ahead for the defense industry, as the trends suggest, depends on how the entire industry maximizes the potential of AI executing the right strategies, tools and avoiding common AI investment mistakes. But first, how far has AI gone in modern military tactics?MORE FOR YOUAI in Modern Military TacticsFrom autonomous drones to predictive analytics and real-time intelligence processing, AI has enhanced efficiency and precision on the battlefield. The spectral battlefield which is defined in the Joint Doctrine Note 3-16 as the range of all electromagnetic radiation is a great example of how AI is quickly transforming the defense space. Put simply, the modern spectral battlefield is the military environment where the control of the electromagnetic spectrum, from radio or radar frequencies to wireless cyberspace applications to signal intelligence and more, are controlled or contested.AI-driven signal intelligence and management platforms, Friedland noted, have become indispensable for dominating the increasingly complex electromagnetic spectrum. AI is changing how defense organizations monitor and control the spectrum, providing persistent insights and enabling rapid responses to new signal forms, anomalies and threats that impact the operation of friendly systems and electronic countermeasures, even in contested environments.For example, Sensorzs SWORD platform leverages cutting-edge AI to analyze unprecedented quantities of spectrum data in real time and automatically classify wireless transmissions. On the battlefield, where every second counts, AI delivers a critical advantage to fighting forces, said Friedland.A great example of where this recently happened is Ukraine, where AI-powered tools have been used to intercept and analyze enemy communications, offering real-time insights that have proven invaluable in tactical decision-making.Crafting Effective Strategies for AI in DefenseBuilding a successful AI strategy begins with aligning technological goals with operational needs. For defense organizations, this means designing systems that address specific challenges, from real-time battlefield decision-making to supply chain optimization. Friedland said that developing internal expertise in areas like machine learning, data annotation and MLOps is critical. Without these capabilities, organizations will struggle to evaluate AI systems during acquisition or effectively implement them in operations.Equally important is embracing open architecture environments, which enable defense departments to rapidly integrate cutting-edge technologies and ensure compatibility across various AI solutions. Open systems allow for innovation, reduce dependency on single suppliers, and ensure adaptability in fast-changing conflict scenarios, Friedland told me in an interview.AI strategies must also focus on practical technologies like edge computing and multi-sensor fusion. Edge-optimized AI is vital for autonomous drones operating in remote areas, while multi-sensor fusion integrates data from diverse sources to enhance real-time decision-making. When combined, these approaches could ensure AI solutions are more effective even in resource-constrained or unpredictable environments.AI Investment Mistakes In DefenseAs powerful as AI is, implementing it the wrong way could be even more catastrophic than not having it in place. One major mistake that any defense agency can make, according to Friedland, is neglecting the quality and integrity of data used to train AI systems. This is a major issue, which Ive covered extensively in the past.Flawed or biased datasets can lead to disastrous outcomes, such as misidentifying threats or causing collateral damage. Robust data governance and validation are essential to ensure the reliability and fairness of AI models, Friedland noted.Cybersecurity vulnerabilities present another major challenge. AI systems are particularly susceptible to attacks like data poisoning and model manipulation, where adversaries intentionally compromise training data or exploit weaknesses in the system. Defense organizations must implement adversarial testing and robust cybersecurity protocols to counter these risks.A third pitfall is the overhype surrounding AIs capabilities. Defense departments sometimes overestimate AIs potential to solve complex challenges without human oversight. As Friedland warned, AI is a tool to augment human decision-making, not replace it. Critical operations must always involve human-in-the-loop systems to maintain accountability and operational integrity. Unrealistic expectations not only risk operational failures but also erode trust in the technology.A Future Defined by Adaptability and CollaborationAs defense departments accelerate AI adoption, the technologys influence on warfare and geopolitics will only grow. Smarter, decentralized systems like microgrids are emerging as a solution to ensure energy reliability in remote bases, while dual-use technologies applicable to both civilian and military sectors are driving innovation across industries. These developments, however, bring regulatory and ethical challenges, particularly as AI capabilities outpace existing frameworks.In Friedlands view, though, the future of AI in defense will be defined by adaptability and collaboration. The organizations that succeed will be those that invest in future-proof systems capable of evolving with technological advancements and operational demands, he said. This includes advancements in AI-driven predictive analytics, enhanced edge computing, and scalable solutions for integrating AI into multi-domain operations.Ultimately, addressing the challenges of AI adoption in defense will require more than technological breakthroughs. It will demand strategic vision, robust safeguards and ongoing collaboration among governments, industry leaders and innovators.It's also critical, Friedland noted, for defense organizations and their AI system developers to focus on the human factor, ensuring that the necessary expertise and resources are in place to effectively support AI throughout its entire lifecycle.By aligning innovation with collaboration and responsibility, we can unlock the full potential of AI to strengthen defense capabilities and safeguard global security, he concluded.