Ethical AI governance in defense is stuck in a paradox: principles are everywhere, but implementation is nearly nowhere. A sweeping systematic review of 1,085 publications reveals why frameworks designed to keep artificial intelligence safe and accountable in military contexts consistently fail to translate from theory into practice, even as the stakes grow higher.

The research, published in June 2026, exposes a troubling reality: as AI integrates deeper into defense systems, the risks multiply. Ungoverned AI in military applications could escalate conflicts, violate international humanitarian law, undermine protections for non-combatants, and destabilize the international order. Yet despite a decade of research identifying ethical principles that should guide AI in defense, concrete, enforceable governance mechanisms remain largely absent.

What Are the Seven Barriers Blocking AI Ethics in Defense?

The research identifies seven interconnected barriers that operate as a system, each reinforcing the others. Understanding these barriers is critical because they reveal why piecemeal fixes often fail; addressing one problem in isolation simply shifts the failure elsewhere in the governance structure.

• Governance and Structural Barriers: Lack of authoritative, interoperable governance architectures that can transcend organizational and national boundaries, making coordination across institutions nearly impossible.
• Conceptual Barriers: Disagreements over fundamental definitions and frameworks for what ethical AI governance should actually mean in practice.
• Strategic Barriers: Misalignment between ethical governance goals and broader military strategy, operational priorities, and geopolitical competition.
• Operational Barriers: Practical challenges in implementing ethical governance throughout the AI lifecycle, including insufficient oversight of autonomous systems.
• Relational and Cultural Barriers: Organizational cultures that prioritize speed and capability over ethical considerations, plus poor communication between teams.
• Technical and Data Barriers: Algorithmic failures, insufficient data quality, and technical limitations that make ethical governance difficult to enforce.
• Resource Constraints: Insufficient funding, staffing, and interdisciplinary expertise dedicated to ethical governance implementation.

Interestingly, governance and structural barriers dominate the academic literature, but this prominence may reflect their visibility rather than their actual importance. Technical, operational, and cultural dynamics are harder to trace but prove equally consequential in determining whether ethical governance actually works.

Why Does Implementation Fail Where It Matters Most?

The research reveals a critical insight: failure is systemic, not isolated. When organizations attempt local fixes to address one barrier, they often inadvertently create problems elsewhere in the governance system. This interconnected nature means that incremental improvements rarely solve the underlying problem. Instead, the barriers reinforce each other, creating a self-perpetuating cycle where ethical governance remains aspirational rather than operational.

The stakes in defense are uniquely high. Unlike other sectors where AI errors might reduce efficiency or cause financial loss, algorithmic failures in military contexts can cost lives, undermine the protection of civilians, and destabilize international security. Yet this urgency has not translated into effective governance mechanisms. The research notes that AI adoption in defense is advancing rapidly and without concrete, enforceable governance mechanisms, even as high-level ethical principles continue to proliferate.

How Can Organizations Overcome These Barriers?

The researchers propose three strategic interventions that address the systemic nature of the problem rather than treating barriers in isolation. These meta-interventions recognize that governance complexity requires coordinated action across multiple levels.

• Establish Authoritative Governance Architectures: Create interoperable governance structures that transcend organizational and national boundaries, enabling coordination across defense institutions and countries rather than allowing each to operate independently.
• Build Institutional Capacity: Invest dedicated resources and recruit interdisciplinary expertise, recognizing that ethical governance requires specialists in ethics, technology, policy, and military operations working together.
• Integrate Ethics Throughout the AI Lifecycle: Embed ethical governance as a core capability from the earliest stages of AI development rather than treating it as a compliance retrofit added at the end.

This systemic approach represents a fundamental shift in how organizations think about ethical governance. Rather than creating separate ethics committees or compliance teams, the research suggests that ethical considerations must be woven into every decision, from initial design through deployment and ongoing monitoring.

The implications extend beyond defense. The barriers identified in military AI governance are not unique to that domain. Similar structural, conceptual, and resource challenges obstruct ethical governance in other high-risk sectors including healthcare, criminal justice, and financial services. Understanding why implementation fails in defense provides a roadmap for improving governance across these critical domains.

As AI capabilities accelerate and military applications expand, the gap between ethical principles and operational reality grows more dangerous. This research provides the first comprehensive mapping of why that gap exists and offers concrete pathways toward closing it, but only if organizations are willing to undertake the systemic reforms the evidence demands.