AI in Warfare: Who Controls the Kill Chain?

Overview

The integration of artificial intelligence into military operations is no longer a theoretical prospect; it is a present reality reshaping modern conflict. From the trenches of Ukraine to the vast maritime approaches of the Indo-Pacific, autonomous systems are demonstrating that speed and scale can outweigh traditional tactical advantages. However, this technological leap forces a critical question: who ultimately controls the kill chain when machines can detect, decide, and strike with minimal human intervention?

Recent analysis from strategic experts highlights a growing tension between the operational necessity of rapid AI-driven decisions and the legal, ethical, and coalition-based requirements for human oversight. As nations race to deploy uncrewed platforms, the definition of "control" is shifting from direct piloting to algorithmic supervision, raising profound concerns about accountability and interoperability among allies.

The Battlefield Reality: Ukraine and the Autonomous Arms Race

The Russia–Ukraine war has become the primary testing ground for AI-powered autonomous weapons. These systems offer distinct advantages in speed, scale, and cost-efficiency, but they also introduce significant risks regarding command and control.

David Kirichenko observes that autonomous drones are already entrenched in the front lines, with both nations engaged in a rapid arms race to enhance their capabilities. Ukraine has equipped its uncrewed ground vehicles with AI targeting systems, allowing them to identify enemy forces from greater distances with improved accuracy. Furthermore, mid-range drones such as the Hornet, Darts, and RAM-2X utilize autonomous terminal lock guidance, making them more resilient against electromagnetic warfare and jamming.

Ukraine is equipping its uncrewed ground vehicles with AI-powered targeting, letting them spot enemy forces from farther away and more accurately.

Russia is responding in kind. Its Geran Seeker and Molniya strike drones now feature mesh networking and onboard AI for autonomous terminal guidance, and these systems are widely deployed. Kirichenko notes that "robot versus robot battles are common" in this theater. Ukraine’s strategy involves deploying autonomous robots to clear enemy positions before troop advances, while naval drones lie in wait to be activated remotely before striking autonomously.

In long-range strikes, AI fuses battlefield data to identify targets and calculate optimal flight paths, significantly increasing the success rate of deep strikes inside Russian territory. This evolution suggests a future where human soldiers are supported, rather than leading, the initial phases of combat.

Strategic Shifts in the Indo-Pacific

The lessons from Ukraine are reverberating through the Indo-Pacific, prompting defense establishments to reconsider their reliance on traditional, high-cost platforms. Malcolm Davis argues that the Australian Defence Force’s (ADF) dependence on small numbers of costly crewed assets is increasingly out of step with the rapid innovation cycles of uncrewed systems.

Rather than pursuing full autonomy, Davis advocates for semi-autonomous systems that can be acquired in volume and at low cost. This approach aims to build "combat mass" across the region’s vast distances, a crucial factor in deterrence. To keep pace with fast-moving operations where communications may be severed, the ADF may need to shift its doctrine from keeping humans "in the loop" to keeping them "on the loop."

Rather than remotely piloting drones, operators would only oversee what the machines do, leaving more decisions to AI.

This model reduces the cognitive load on operators, allowing AI to handle tactical decisions while humans monitor for anomalies or ethical breaches. The implications extend underwater as well. Through the AUKUS partnership, Australia, Britain, and the United States are developing autonomous submarines like Anduril’s Ghost Shark. If the 2027 delivery target is met, these systems will provide critical protection for subsea infrastructure, such as the internet cables that carry nearly all of Australia’s traffic.

The Governance Gap and Automation Bias

Despite the rapid deployment of these technologies, governance has struggled to keep pace. Aina Turillazzi points out that while AI is financed across almost every line of the Australian defense budget, it is not governed with the same comprehensiveness. She warns that without deliberate pauses for human review at key escalation points, the people meant to oversee AI systems risk becoming little more than a rubber stamp.

The danger lies in automation bias, a psychological tendency where operators under pressure defer to AI outputs. Because AI systems are probabilistic and often opaque, humans may drift from assessing evidence to simply confirming recommendations. This dynamic conflicts with the legal requirement for informed human judgment in targeting decisions.

AI is now financed across almost every line of the Australian defence budget. It is not, however, governed across them.

Turillazzi urges Canberra to build in safeguards, audit AI’s role in high-stakes choices, and publicly commit that AI will not have final authority over the most consequential decisions. Military AI does not need to become conscious to outpace human control; it only needs to be fast enough that human reaction becomes a liability.

Coalition Interoperability and Legal Fragmentation

Even if individual nations establish robust rules, international cooperation faces significant hurdles. Samuel White raises the issue of differing legal and operational frameworks among partners. In a hypothetical coalition operation, one nation might act on an AI’s targeting recommendation, while another demands human verification of the underlying intelligence, and a third requires the AI to explain its logic.

Although AI is intended to accelerate decision-making, incompatible approaches to law and human judgement can slow a coalition down or, worse, produce disagreement only after a decision has been made.

Such fragmentation could paralyze joint operations or lead to catastrophic misalignment. The challenge is compounded by continual learning, where AI systems update their behavior with new data post-deployment. As Ayeza Areej warns, this adaptability makes behavior harder to predict, especially when adversaries use deception or electronic warfare to feed misleading data.

As the kill chain becomes increasingly automated, the military community must ensure that the speed of algorithmic decision-making does not outstrip the capacity for human responsibility and international legal consistency. The future of warfare will be defined not just by the sophistication of the AI, but by the clarity of the human rules governing it.