In their September 2024 op-ed essay in The New York Times, Raj M. Shah and Christopher M. Kirchhoff argue that the United States must hastily embrace artificial intelligence (AI)-powered weapons systems to remain competitive in the “civilisational race” against China. They call for an overhaul of the U.S. military’s technological capabilities to keep pace with AI-driven warfare. And where the US go, the UK follows.
It has recently been reported that the British Army will soon be using AI-powered robots, known as SimStriker, to train recruits in battlefield scenarios. These robots can react to soldiers’ movements and spoken commands, and they are capable of firing BB pellets as part of the training simulation. This new technology aims to provide more realistic training experiences for soldiers by simulating hostile and dynamic combat conditions. The deployment of AI in such training scenarios represents a shift towards more advanced, tech-driven methods in military preparation. From there, it’s use in the battlefield seems an almost certainty.
But this enthusiasm for an AI arms race, without proper consideration of the profound ethical and humanitarian risks, is dangerous and misguided.
Shah and Kirchhoff paint AI as the next revolution in military affairs, likening it to the technological leap the Maxim gun provided during the First Matabele War. Yet their historical framing ignores one critical reality: weapons, no matter how advanced, can cause disproportionate suffering when deployed without the necessary legal and ethical safeguards. The authors’ call to reduce human involvement in decision-making, shortening the kill chain with autonomous systems, fails to reckon with the well-documented risks AI weapons pose to civilian lives and international law.
As Dr. Iain Overton, Executive Director of Action on Armed Violence, argues: “The notion that AI weaponry can operate with precision is a dangerous myth—these systems are prone to malfunction, can hallucinate, can – with poor programming – cause indiscriminate civilian harm, and can, in the end, undermine the foundational laws of war.”
Autonomous weapons, by their very nature, remove the human judgement that international humanitarian law (IHL) demands in conflict zones. The authors in the NY Times optimistically describe AI as more efficient and accurate, but this ignores the risks inherent in letting machines make life-and-death decisions. The potential for AI misidentifying targets, being hacked, or malfunctioning is immense, as is the threat of mission creep—where nations, emboldened by these technologies, engage in riskier military operations that could lead to more conflict, not less.
We’ve seen this before with so-called “miracle weapons.”
The F-35 fighter jet, for instance, was heralded as a game-changing technology. Yet two decades later, it remains plagued with performance issues and has become the most expensive weapons program in U.S. history. The history of such systems, littered with cost overruns and underperformance, raises serious doubts about the claims that AI systems will somehow be cheaper, faster, and more reliable. As always, military contractors and venture capitalists are the ones set to benefit, while the public faces the consequences of inflated defence spending and increased risks of war.
You can hear the defence companies salivating at the prospects of very human profits in this rush to battlefield robotics.
Shah and Kirchhoff’s essay also plays into a dangerous, techno-enthusiastic narrative by positioning the U.S. in a “civilisational race” with China. This framing prioritises geopolitical rivalry over the real human costs of war. The authors conveniently gloss over the fact that pushing for faster deployment of AI weapons increases the chances of civilian casualties, particularly in asymmetric conflicts where AI cannot easily adequately distinguish between combatants and non-combatants.
The fact that Shah, now a managing partner at Shield Capital, and Kirchhoff, who helped build the Pentagon’s Defense Innovation Unit, could even profit from accelerating the AI arms race should give us pause. Their essay reads more like a sales pitch for the military-industrial complex than a serious reflection on the risks posed by these emerging technologies.
AOAV contends that the New York Times should have applied a more critical lens to this issue, given the enormous stakes for global security.
Rather, instead of racing ahead with AI weaponry, the U.S. should focus on creating stringent legal frameworks to control the use of autonomous systems in warfare. The notion that AI can replace human judgement in matters of life and death not only violates the principles of IHL but also sets a dangerous precedent for how conflicts will be waged in the future.
We should be deeply sceptical of any argument that suggests we remove human accountability from warfare, especially when lives—civilian or otherwise—are at stake.
So while innovation in military technology may be inevitable, it must not come at the expense of humanity and morality. The New York Times missed the mark by publishing an uncritical essay that pushes us closer to an unchecked AI arms race.
Rather than accelerating toward a future of autonomous warfare, we must prioritise global cooperation and transparency, ensuring that technology is used to prevent conflict, not perpetuate it.
Did you find this story interesting? Please support AOAV's work and donate.
Donate