New reports from international leaders and military ethicists highlight a critical failure to protect human agency as automated systems assume control over high-stakes decisions.

Global governance frameworks and military operational standards are currently failing to protect human agency as artificial intelligence assumes greater control over high-stakes decision-making. Two distinct reports released in September 2026—one from the international organization known as The Elders and another from a military conference involving US Army chaplains—converge on the finding that current oversight mechanisms are insufficient to manage the ethical and existential risks of automated systems.

The Elders, a group of former heads of state and government, published ten principles on September 21, 2026, characterizing the current state of AI governance as a systemic failure of sovereign responsibility. This document argues that governments have largely abdicated their duty to regulate AI, allowing private entities to dictate the terms of public safety, human rights, and the use of force. The Elders assert that AI governance is a fundamental obligation of the state rather than a voluntary practice for technology developers.

The concentration of AI development within a small cohort of private firms, including OpenAI, Anthropic, Google, Meta, Microsoft, DeepSeek, and Alibaba, has created a market environment where oversight is virtually nonexistent. The Elders emphasize that AI capabilities frequently transcend national borders, necessitating immediate negotiations between the United States and China to establish international safety standards. These standards would aim to enforce existing laws and curb the excessive power currently held by a handful of technology corporations.

Military institutions are simultaneously grappling with the integration of AI into combat operations, as evidenced by a September 18 conference held by Task & Purpose within US Army Europe and Africa. The deployment of AI-enabled sensor-to-shooter artillery processes, such as those tested by the 25th Infantry Division, has moved beyond the hypothetical stage. This integration forces service members to confront the moral weight of lethal actions that are increasingly guided by algorithmic targeting data.

The Elders argue that the principle of restraint must be applied to all decisions involving legitimate force, rejecting the use of weapons that select their own targets. While current Pentagon policy mandates human judgment over autonomous weapons, the practical reality of high-speed combat often undermines this requirement. Meaningful control requires that operators have the authority and the technical literacy to inspect, contest, or reverse machine-generated outputs during critical operations.

The Elders’ report further clarifies that leaders cannot effectively supervise systems they cannot examine, question, or stop. This requirement for transparency extends to the data supporting these models and the distribution of benefits or losses resulting from their deployment. By framing these needs as obligations, the organization seeks to shift the focus from voluntary corporate ethics to mandatory sovereign oversight.

Jesse Kirkpatrick, a research professor at George Mason University, notes that the psychological architecture of modern combat is evolving faster than the pastoral care infrastructure intended to support soldiers. He warns that moral injury—a form of psychological harm resulting from actions that violate deeply held beliefs—is a growing risk for troops who possess nominal responsibility for lethal outcomes without genuine control over the underlying AI systems. The potential for moral injury arises when operators are pressured to approve machine-generated recommendations without sufficient time or information to challenge them.

Effective AI governance requires a shift from viewing systems as static tools to treating them as dynamic, evolving risks that demand continuous monitoring. Organizations must track errors, security vulnerabilities, and unintended behavioral effects long after a system is deployed. This strategic visibility must bridge the gap between technical performance metrics and the human experience of those who rely on these automated recommendations.

The path forward necessitates a move toward enforceable international rules and transparent institutional frameworks that prioritize human safety over corporate profit. Decision-makers in both the boardroom and the military command structure must ensure that systems remain subject to human intervention at every stage of the decision-making process. Without such safeguards, the reliance on opaque algorithms threatens to erode the accountability structures that underpin both democratic governance and ethical military conduct.

At Global Business & Economics Journal, we contribute to the  future of business and technology with news,  innovative research, expert insights, and thought-provoking analysis.