As artificial intelligence (AI) rapidly integrates into various aspects of modern life, its application in military decision-making is drawing significant scrutiny from lawmakers. Concerns are mounting within the United States Congress regarding the potential for AI to autonomously generate target lists, formulate operational plans, and influence critical combat decisions without sufficient human oversight or transparency. In response, a key congressional committee is moving to expand existing limits on autonomous weapons to encompass these sophisticated AI-driven planning functions.
The House Armed Services Committee’s (HASC) proposed version of the 2027 National Defense Authorization Act (NDAA) includes a directive for the Pentagon to revise Defense Directive 3000.09, a foundational policy document governing autonomous weapon systems. The proposed amendment aims to broaden the directive’s scope to explicitly cover mission planning activities and other AI-enabled systems that materially influence the employment of force.
Should this measure pass, it would represent a “big change” in how the U.S. military governs AI, according to Bryan Clark, a senior fellow and director of the Center for Defense Concepts and Technology at the Hudson Institute. The initiative is part of a broader legislative effort in Congress to address the implications of emerging AI and related advanced capabilities, including proposals for orbital data centers and the establishment of a new combatant command dedicated to overseeing autonomous and robotic systems.
Under the plan inserted into the House markup of the defense bill, the Pentagon would be mandated to establish or revise guidelines for “artificial intelligence-enabled systems intended to support, recommend, or materially influence operational decisions associated with the employment of force,” as detailed by the HASC Subcommittee on Cyber, Information Technologies, and Innovation. This expanded purview would specifically include systems used for “operational planning, target development, weaponeering, or engagement recommendation.”
Defense Directive 3000.09, originally published in 2012 and last updated in 2023, has primarily focused on the parameters for human operators when making target selection and engagement decisions involving autonomous weapon systems. A 2025 Congressional Research Service Report, “U.S. Policy on Lethal Autonomous Weapon Systems,” highlighted the directive’s emphasis on ensuring appropriate levels of human judgment and control in the deployment of such systems. However, the current directive has not explicitly addressed the use of AI in the earlier, upstream phases of military decision-making, such as strategic planning or target identification.
The urgency for such revisions has been underscored by recent military engagements. “The [department] has been using a lot of AI-decision enabled support systems in the war in Iran for strike planning, for example, and what this is saying is that you now need to have some kind of governance around those too,” Clark noted, referring to the recently concluded “Operation Epic Fury.” He elaborated on the evolving complexity of AI’s role: planners are no longer simply “proposing a kill chain to go blow something up, which is pretty simple.” Instead, AI is now being tasked to “decide what are the right things to attack those targets and then what are the things I should use to attack those targets and then build me some kill chains to make that happen.”
Clark highlighted the inherent challenges of this shift. Conventional military planning processes involve numerous human planners with diverse expertise, creating an intricate system with built-in checks and balances. In contrast, while AI systems may account for a multitude of factors, the mechanism by which they arrive at a particular decision often remains opaque—a figurative “black box.” Understanding the rationale behind an AI-generated recommendation, or tracing potential biases in its data or algorithms, becomes exceedingly difficult, posing significant accountability and trust issues.
Retired Air Force Lt. Gen. Bill Bender, a former Air Force Chief Information Officer, echoed the sentiment regarding the value of congressional oversight as AI technology continues its rapid evolution. “Using AI in mission and target planning systems ‘begs for legislative oversight, at a minimum’,” Bender stated. He stressed the importance of a collaborative effort between the Pentagon and Congress to develop a “measured, thoughtful approach” for implementing AI tools. This approach, he argued, must carefully balance the need for robust controls with the imperative not to unnecessarily constrain AI innovation, which could lead to missed strategic opportunities.
The challenge, Bender concluded, “will be loosening the restrictions in a timely fashion as better AI enablement has iteratively arrived upon improvements.” This underscores the dynamic nature of AI development and the need for policy frameworks that can adapt alongside technological advancements, ensuring both ethical deployment and strategic advantage.
Why This Matters
The debate within the U.S. Congress over regulating artificial intelligence in military planning and target selection carries profound implications for global security, ethical warfare, and the future of armed conflict. The proposed expansion of Defense Directive 3000.09 represents a critical pivot from merely controlling autonomous weapons that execute actions to governing the intelligent systems that recommend or even formulate those actions. This shift directly addresses the “black box” dilemma—where AI makes complex decisions without transparent human-understandable reasoning—raising fundamental questions about accountability, responsibility, and human control in lethal operations.
Ethically, the move challenges traditional notions of human moral agency in warfare. If AI systems are recommending targets or operational plans, who is ultimately responsible for potential errors, unintended civilian casualties, or escalations? The lack of full transparency in AI decision-making could complicate investigations into war crimes or ethical breaches, potentially eroding international humanitarian law. Moreover, biased data fed into AI algorithms could perpetuate or exacerbate existing inequalities or conflicts, leading to disproportionate impacts on certain populations.
Strategically, the integration of AI into military planning promises enhanced speed, efficiency, and the ability to process vast amounts of data, potentially offering a decisive advantage in conflict. However, it also introduces new risks: the potential for algorithmic errors to lead to catastrophic miscalculations, vulnerability to sophisticated cyberattacks that could manipulate AI systems, and the risk of an AI arms race among global powers. The push for governance reflects a recognition that unchecked AI development in this domain could destabilize international relations and heighten the probability of conflict. Ensuring human-in-the-loop control for critical decisions becomes paramount to prevent unintended escalation and maintain strategic stability.
For global citizens, this legislative effort highlights the growing intersection of advanced technology and governance. The decisions made today regarding AI in military applications will shape the nature of future warfare, influencing everything from peace and security to the very definition of human autonomy in decision-making. It underscores the urgent need for international dialogue, ethical frameworks, and robust civilian oversight to ensure that powerful AI technologies serve humanity responsibly, rather than becoming instruments of unpredictable and potentially uncontrollable conflict.