Artificial Intelligence and the Urban Battle

Derided for his acting, Arnold Schwarzenegger, nevertheless, made two classic action movies in his career: Terminator and Predator. Terminator, released in 1984, begins memorably with an apocalyptical vision of Los Angeles in 2024. Drones fly overhead firing lasers at fleeing humans, while monstrous robotic vehicles crush human skulls beneath their tracks. The film was a work of science fiction. 

Yet, some commentators today have claimed that the future of urban warfare might approach this cinematic portend. They believe that in the future, the urban battle will be fought not by humans but rather by robots and autonomous weapon systems directed by a super-intelligent computer. Some deplore this prospect– while others embrace it with millenarian enthusiasm, believing the machines will liberate humanity from the horror and waste of war. How accurate are these prognostications of urban warfare? Will the urban battle of the near future really look like those opening frames of Terminator?

It is worth noting that we have actually been in this situation before. In the 1920s and 1930s, many prominent military scientists made dire predictions about the future of warfare. Most famously, Guilio Douhet, Basil Liddell Hart and HG Wells all claimed that strategic airpower – the bomber – was about to revolutionise warfare. On the assumption that "the bomber would always get through", they claimed that any future war would involve the immediate and total destruction of cities. The civilian population, which in Europe had escaped almost totally in the First World War, would become the target, and indeed the victim of an ineluctable new weapon: ‘There is no reason why within a few hours or at most days from the commencement of hostilities, the nerve system of a country inferior in air power should not be paralysed. Imagine for a moment London, Manchester, Birmingham, and half a dozen other great centres simultaneously attacked, the business localities and Fleet Street wrecked, Whitehall in a heap of ruins, the slum district maddened into an impulse to break loose and maraud’.

Yet, contrary to these jeremiads, air forces found it very difficult to conduct effective strategic bombing campaigns. Bombers were frequently shot down; it was very difficult to hit cities – never mind the targets in them. The 1942 Butt Report on the RAF’s campaign against Germany recorded that only one in five RAF bombs fell within five miles of their targets (less in the Ruhr). Some nights, German defenders could not work out what the RAF were trying to hit. Of course, under Air Chief Marshal Arthur Harris, the RAF became far more effective, partly because of the introduction of disruptive technologies: computer-assisted radar navigation and radar-blocking equipment. After the German air defences collapsed in late 1944 and early 1945, the RAF and the USAAF inflicted terrible damage on German cities. Yet, strategic bombing failed to live up to the pre-war hopes or fears. Even with the vast investment in the RAF, strategic bombing proved difficult to do and changed warfare far less than was predicted. States, armed forces and citizens adapted and, while the bombing was certainly regarded as terrible during the Second World War, it became a normal part of war; it was little different from the routine sacking of cities in antiquity and the middle ages.

The example of strategic bombing in the 1940s and 1950s is useful in thinking about the effect of disruptive technologies on urban warfare today. The current debates about AI reprise those discussions of the inter-war period. Now, as then, many scholars and practitioners claim we are at the dawn of an urban revolution. The complex of disruptive technologies – AI, remote and autonomous weapon systems and robotics – will transform urban combat. Cities lie defenceless before these new weapons. An advanced military force of the future will be able to identify targets in a city with absolute fidelity and strike every one with complete precision from remote or autonomous platforms. The urban battle will have been ceded to the machines, which are quicker, more accurate and more lethal than any human could hope to be. 

If the Second World War is taken as a comparator, the predictions are vastly overstated. Although military forces will eventually utilise disruptive technologies in the next decade, their introduction and application will be difficult, expensive and time-consuming. Moreover, disruptive technologies – just like the computer-aided Lancaster Bomber flying over Germany – will have less impact on war than is presumed. Urban warfare comprises three elementary factors: the demography of cities in which the battle takes place, the size and structure of the armed forces that do the fighting, and, finally, the weapons they use. Weapons are extremely important; yet, they are only one part of the urban battle. 

In order to understand the urban battle, it is necessary to consider the other two factors alongside weaponry. Once this urban triad is appreciated, it is possible to recognise the distinctiveness of the urban battle of the twenty-first century. Given the sophistication of the weaponry, recent urban battles have, perhaps surprisingly, been characterised not by surgical strikes or frictionless swarms; they have been defined by the siege. In Mosul, Aleppo, Marawi, and Donetsk combatants engaged in long, slow, attritional wars of geographic position. Opponents converged on decisive locations inside urban areas over which they fought bitterly. In some ways, recent battles have looked more like the gruelling sieges of antiquity rather than the postmodern, virtual engagement of military imagination. 

The question is will AI and disruptive technologies transform the urban battlescape – the inner urban micro-siege — that has characterised the last two decades? There is little doubt that AI and disruptive technologies will have an effect on military operations in the next two decades. Drones have already changed the way armed forces operate. The Azerbaijani army used them to great effect in the recent Nagorno-Karabakh conflict. The Royal Marines are currently exploiting drones to provide intelligence screens and logistic support for their ‘Future Commando Force’. The Royal Tank Regiment has already experimented with robotic military vehicles tethered to Challenger tanks. Similarly, AI offers considerable potential in terms of intelligence gathering and fusion. The aspiration is to collate big data on enemy activity in any theatre from a suite of sensors to improve situational awareness, accelerate responses, and increase the lethality of counter-strikes.

Yet, AI and drones will work in urban battle only if there are still troops on the ground, capable of holding ground, engaging with human populations and executing the missions in conditions that will defy machine intelligence with their complexity and ambiguity. Rather than revolutionise warfare itself, disruptive technologies are more likely to accentuate current trends towards the siege. Cities are vast and expanding. Military forces are small and getting even smaller; the British Army will have only 72,000 regular soldiers by 2025. Urban operations will be very difficult, then. It will be imperative that casualties are minimised because the Army cannot afford to lose many soldiers. Therefore, AI, drones, and autonomous systems will augment smaller, more expert human combat teams, but it is unlikely they will displace soldiers or change the fundamental geometries of urban combat. They will improve the survivability of forces in the urban environment and improve their performance. Yet, on high-intensity operations, those combat teams will be defending or attacking formidably fortified urban positions against cunning and clever human opponents who will always seek to offset technologies, no matter how advanced. It is difficult to see how the fundamental conditions of urban warfare will radically change in the next two decades. Indeed, disruptive technology is likely to intensify the siege conditions of contemporary urban warfare, especially against peer opponents, as each side exploits the potential of AI and remote systems. Urban warfare in the future is likely to look more like ancient sieges than anything we might see in a Schwarzenegger film, no matter how compelling.

Anthony King’s Urban Warfare in the Twenty-First Century (Polity) was published on 23 July 2021.

https://www.amazon.co.uk/Urban-Warfare-Twenty-First-Century-Anthony/dp/150954366X