Artificial intelligence is rapidly becoming the decisive technology of future conflicts reshaping how wars are planned, fought, and won

Introduction: The Technology That Changes Everything

Every generation of military technology has produced a period of strategic uncertainty — the era between the emergence of a new capability and the development of the doctrines, countermeasures, and international frameworks adequate to manage it. The introduction of air power in the early twentieth century, nuclear weapons in the mid-twentieth century, and precision-guided munitions in the late twentieth century each produced such periods of adjustment, sometimes at enormous cost.

Artificial intelligence is now producing the same kind of strategic disruption — more rapidly, more pervasively, and across a broader range of military applications than any previous technology transition. From autonomous drones that can identify and engage targets without human control, to intelligence analysis systems that can process satellite imagery faster than any human analyst, to cyber warfare tools that can probe and exploit network vulnerabilities at machine speed, AI is being integrated into military systems at a pace that existing governance frameworks, operational doctrines, and strategic concepts were not designed to accommodate.

The race to integrate AI most effectively into military operations is now explicitly framed as a strategic priority by the United States, China, and major European powers. The outcome of that competition — which country builds the most capable and most reliable AI-enabled military, and which establishes the doctrinal and operational frameworks to use it most effectively — may be as consequential for twenty-first century strategic balance as the nuclear competition was for the twentieth.

Section I: What Military AI Actually Does

Intelligence, Surveillance, and Reconnaissance

The application of AI to military intelligence analysis is among the most mature and most impactful of all current military AI deployments. Satellite and aerial surveillance now generates imagery data at volumes that dwarf human analysts’ capacity to review it in a timely fashion. AI systems trained to identify specific objects — military vehicles, construction activity, weapons systems, personnel movements — can process this imagery automatically, flagging items of interest for human review at speeds and scales that have fundamentally changed the intelligence cycle.

The practical military significance is substantial. Activity that might previously have taken days to identify through manual review can now be detected within hours. The ability to track military movements across large territories at near-real-time speed provides a strategic intelligence advantage that was previously achievable only with very large analyst workforces — and is now achievable with relatively modest AI infrastructure.

Targeting and Precision Fires

AI systems are being integrated into targeting processes for both conventional and precision strike weapons. Applications range from assisted target recognition — AI systems that help operators identify targets in imagery or sensor data — to more autonomous systems capable of developing targeting solutions with varying degrees of human oversight. The Ukraine conflict has provided operational experience with AI-assisted artillery targeting, drone guidance, and electronic warfare applications that is directly influencing procurement and doctrine across multiple militaries.

Autonomous Systems and Unmanned Warfare

The integration of AI with drone systems is producing the most visible and strategically consequential current application of military AI. Drones with varying degrees of autonomous target selection and engagement capability have been deployed or are in development across multiple countries. The Ukraine conflict has demonstrated the operational effectiveness of large-scale drone operations. The next generation of AI-enabled drone systems is expected to operate at greater range, with reduced human control loop requirements, and in swarm configurations that overwhelm conventional air defence through coordination and numbers.

Section II: The US-China AI Competition

The United States and China are engaged in the most significant military technology competition since the Cold War nuclear arms race — and both governments explicitly identify AI as its central domain. The United States maintains advantages in AI research depth, semiconductor design capability, and the talent base concentrated in its leading AI research institutions and technology companies. China has made AI a national strategic priority, with the 2017 ‘New Generation Artificial Intelligence Development Plan’ explicitly targeting AI leadership by 2030 and integrating AI development into its military modernisation programme as a central objective.

The competition operates simultaneously in research, in commercial AI development with military applications, in semiconductor access, and in the recruitment of AI talent. American export controls on advanced semiconductors, aimed at limiting China’s access to the computing hardware necessary for training large AI models, represent an attempt to maintain a technological advantage that China is working intensively to close through domestic semiconductor development. The outcome of this competition will have implications extending well beyond the purely military domain — into economic competitiveness, technological standards, and the governance frameworks that will shape AI development globally.

Section III: Governance, Ethics, and the Autonomous Weapons Challenge

The integration of AI into military systems raises governance questions that the international community has not yet resolved. At the centre of these questions is the issue of meaningful human control over lethal decisions — the principle that decisions to use lethal force should retain genuine human judgment rather than being delegated entirely to autonomous systems. This principle is embedded in international humanitarian law through the requirement for distinction between combatants and civilians, proportionality assessment, and precaution in attack — requirements that are difficult to verify in AI systems operating at machine speed.

A campaign by a coalition of non-governmental organisations and academic researchers to negotiate a binding international treaty on autonomous weapons systems — sometimes called ‘killer robots’ — has been under discussion at the United Nations since 2014 without producing a binding agreement. States with significant military AI programmes have been resistant to binding constraints on systems they regard as providing genuine military advantage. This governance vacuum creates risks both of unintended escalation and of norm erosion, as the absence of agreed international standards makes it more difficult to establish shared expectations about the boundaries of acceptable military AI application.

Conclusion: The Algorithm and the Battlefield

The race for AI military dominance is already underway and is accelerating. Its outcome will be determined not only by which country develops the most powerful AI algorithms but by which military can integrate AI most effectively into its operational processes — procurement decisions, training, doctrine, command architecture, and the human-machine teaming that will define how AI-enabled militaries actually function in practice. The countries that succeed in this integration will have genuine and potentially decisive military advantages in the conflicts of the coming decades. The countries that fail to adapt will face a strategic disadvantage as significant as the gap between industrialised and non-industrialised militaries in the nineteenth century.

North Korea’s New Missile Doctrine