Science fiction had predicted it: wars without soldiers

“I’ve seen things you humans couldn’t imagine: battle ships on fire off the bastions of Orion, and I’ve seen B-rays flashing in the dark near the gates of Tannhäuser. And all those moments will be lost in time, like tears in rain. ‘It’s time to die’,” is the famous monologue delivered by Rutger Hauer as the replicant Roy Batty in “Blade Runner”, recalling his experience as a soldier-android in the off-world colonies of space. “Blimey, Valentina, you’ve made a right mess of things! Everyone’s been cut to pieces… I didn’t think you were capable of…” “…of such ruthlessness, you mean? But I haven’t killed anyone! Have you seen what’s inside these armour suits? Nothing! They’re empty! Didn’t you know that soldiers on Earth have now been replaced by robots?”, explains Valentina after she, riding her horse Potëmkin, has slaughtered a host of Giovesan gendarmes – who are, in fact, empty armour suits animated by robots – in “Valentina the Pirate”, the twenty-third comic strip featuring the famous character created by Guido Crepax. Roy Batty is a biological android: ‘More human than human’, according to the slogan of the Tyrell Corporation, which produced him using advanced genetic engineering techniques, even though he is programmed to live for only four years. Conceived in a 1982 film, he dies in a dystopian Los Angeles in 2019. Published in 1976 as a sequel to another story from 1967, “Valentina pirata” is set in the year 2570, with costumes that curiously evoke the 16th century, even though it is in fact a dream; and it features mechanical robots.

A cross between the two – part machine, part biological – is the Terminator played by Arnold Schwarzenegger, who travels back from 2029 to 1984 (the year the film was released) specifically to carry out a rebellion by artificial intelligence against humanity, parallel to that in "Blade Runner": although not exactly the same, and although he later changes sides in the sequels. Not all science fiction is actually like that. In "Starship Troopers", Robert Anson Heinlein’s famous 1960 novel and its equally famous 1997 film adaptation, it is precisely the fact that the war between Earthlings and extraterrestrials is highly technological that requires the indispensable complement of a highly specialised human element, driven by the fact that military service is a prerequisite for enjoying political rights. If you like, this seemed to be, to some extent, the logical evolution of the process of the industrialisation of war that began in 1914. First, the masses of ‘workers of death’, deployed thanks to the technological advances in mobilisation and supply during the First World War. Then came the further development of technological elements to break the deadlock of trench warfare – a process that began in 1914–18 with aircraft and tanks, but culminated in the Second World War, culminating in the atomic bomb. Following this, a further wave of technological advancement led to the general shift from conscription to professional armies: yet, even here, there was a particular emphasis on the essential human complement provided by Special Forces. The Space Infantry of the 21st century.

But then the war in Ukraine broke out. On the one hand, it was like a great return to 1914–18, with masses of people thrown into the trenches and hand-to-hand combat in which blows with a spade once again became a more reliable weapon than bayonets, which can become lodged in the body they pierce: as Erich Maria Remarque had already described in "Niente di nuovo sul fronte occidentale". But, once again as in 1914–18, this has led to a further drive towards the technologisation of war, fulfilling ‘Valentina the Pirate’s’ prophecy five centuries ahead of time: a ‘war without soldiers’ which Ukraine, in particular, is pursuing, given its demographic disadvantage; and which is already raising questions amongst those who have begun to study the phenomenon, even if, for the time being, the fear is not so much of a ‘Terminator’- or ‘Matrix’-style robot uprising, but rather a privatisation of war that could, in some respects, hark back to the mercenary companies of the late Middle Ages; for others, to the 17th-century East India Companies; and for others still, to something that has, in fact, never been seen before.

In reality, the impetus had already come primarily from those nations which, as the two leading military superpowers, also have the greatest incentive to remain at the cutting edge of military technology. In the United States, in particular, on the Pacific Coast, the San Francisco-based start-up Foundation Robotics – founded in 2024 with a name that appears to pay homage to Isaac Asimov by combining his two series on Robots and the Foundation – has created the Phantom MK-1: the first true humanoid designed for defence to be tested in real-world scenarios, such as the rear lines in Ukraine. Standing 1.75 metres tall and weighing 80 kilograms, with five fingers on each hand and a camera in place of its eyes, it is deployed for logistical tasks, ammunition transport and reconnaissance in high-risk areas. Following the first 40 operational units in 2025, plans envisage a further 10,000 by the end of 2026 and 50,000 by 2027. These will not be sold but offered on a leasing basis, at around 100,000 dollars a year. On the Atlantic Coast, meanwhile, is Boston Dynamics, which – with public funding from the Defence Advanced Research Projects Agency – unveiled the bipedal android Atlas back in 2013, initially designed for various search-and-rescue tasks. Since 2024, it has been developed into an electric version suitable for the automotive industry’s assembly lines, and has even been taught to play football. However, it can replace soldiers in transport and reconnaissance tasks in dangerous – or even simply inaccessible – areas.

In China, it is the Shenzhen-based company Ubtech, founded in 2012, which is seeking to mass-produce the Walker S2: an industrial humanoid robot that is approximately 1.76 m tall, weighs 70 kg, can lift up to 15 kg and has a battery life that allows it to replace its own battery independently. All this at a cost of between 68,000 and 100,000 dollars per unit. It was designed primarily to work on factory assembly lines, but the People’s Liberation Army is also trialling it for patrolling borders and assisting law enforcement agencies. In theory, in keeping with the Laws of Robotics devised by Asimov himself – which prevent a robot from harming a human being – these machines are not currently intended to replace combat soldiers, but rather for tasks involving high logistical and operational risk: carrying rucksacks, ammunition and supplies into areas inaccessible to wheeled or tracked vehicles; crossing minefields; inspecting environments contaminated by chemical, biological or radiological agents; exploring and clearing buildings or enclosed rooms where the risk of ambush for human soldiers is highest; and evacuating the wounded from areas under enemy fire.

However, LAWS (Lethal Autonomous Weapons Systems) – capable of deciding to kill without direct human control – are just around the corner. Companies such as Foundation Robotics are currently maintaining a clear line: navigation is autonomous, but any decision relating to the use of force or targeting remains strictly subject to remote control by a human operator. However, when guided by a human, robots are already capable of killing. Moreover, even more so than robots, drones are now a key element in warfare. France, in particular, has been running the Pendragon Project since 2025 to develop both swarms of aerial drones linked to autonomous ground vehicles capable of mapping and striking targets in real time, minimising human presence on the ground; and heavy underwater drones, capable of patrolling the seabed autonomously for months at low cost; as well as a robotic combat unit which, from 2027, will be capable of carrying out some of the tasks of an infantry company or a cavalry squadron, under the remote control of 15 military personnel.

However, these are all programmes currently under development. It was, in particular, in December 2024 that the 13th Khartiya Brigade of the Ukrainian National Guard carried out a combined ground and air assault near the town of Lyptsi, without a single Ukrainian soldier setting foot on the battlefield. Unmanned ground vehicles with machine-gun turrets, aerial drones equipped with grenades and assault rifles, and conventional kamikaze munitions seized Russian positions, cleared mines and established a defensive perimeter. Palladium Magazine, a San Francisco-based publication of the American Governance Foundation – known for its emphasis on elite theory and institutional dynamics – takes this episode as the starting point for an analysis by Juan Severini: a PhD candidate in the Department of Political Science at Georgetown University and co-director of the Initiative for Artificial Intelligence and Democratic Citizenship at the same university. This is, in fact, one of the most striking demonstrations of a new paradigm of warfare in which robotic warfare is establishing itself as a direct response to the shortage of human manpower and the political costs of combat casualties. The growing use of autonomous systems, the article explains, is not merely a tactical innovation but a structural transformation of political power. In contrast to the ‘Space Infantry’ model – which transposed the ancient concept of the citizen-soldier from the Greco-Roman world into the realm of science fiction – the replacement of soldiers by robots effectively shifts control of warfare from public institutions to the private companies that design, programme and operate these machines.

The essay analyses in detail the particularly adverse demographic situation facing Ukraine, whose human resources are being depleted by mass emigration, an ageing population and combat casualties. In this context, even the least efficient robotic soldiers represent a profitable investment if they help to preserve the workforce and ensure a steady supply of recruits. With a greater propensity to mobilise vast reserves and limited access to advanced microelectronics, Russia, by contrast, operates according to the opposite logic: more human capital, less cutting-edge technology. But this difference illustrates a broader tension in contemporary military strategy: the balance between capital and labour. States facing labour shortages tend, in fact, to automate more rapidly; those with an abundance of recruits can afford to postpone this transition. Even an ‘America First’ approach to foreign policy argues that a reduction in US personnel abroad could be largely offset by greater use of unmanned munitions, thereby reducing costs and bringing soldiers home without weakening deterrence. The analysis warns, however, that as full autonomy becomes technically feasible, the absence of humans on the battlefield will become the norm, reducing the political costs associated with starting or prolonging conflicts. In short, the world of Valentina the Pirate.

The way this new dependency works is exemplified by Palantir, whose CEO, Alex Karp, visited Ukrainian President Volodymyr Zelensky shortly after the Russian invasion of 2022. As a result, the company’s computer vision algorithms helped Ukrainian forces track enemy positions in real time during the 2023 counter-offensive. What began as a $70 million contract for Project Maven – originally developed with Google and subsequently expanded to include Anthropic – has grown to $1.5 billion, of which nearly $800 million was carried out by Palantir. But this same technological dependence now extends to the major powers as well. During the recent conflict with Iran, SpaceX and the Pentagon negotiated a ‘subscription tier’ for the control of unmanned munitions, which effectively doubled the cost of each drone at the height of hostilities. But this episode highlights a paradox: the labour savings promised by robotic warfare are negated if the private ecosystem that underpins it sets its own prices in times of war. Elon Musk’s influence over access to the Starlink network – contested by both Ukraine and Russia – is another example cited by Severini to illustrate the extent to which a CEO’s foreign policy can shape the course of an armed conflict. Private companies, free from the wage restrictions or recruitment bureaucracy of the public sector, enjoy a competitive advantage when it comes to the specialised talent required by these technologies, and this advantage translates into bargaining power with states.

We return to those mercenary companies which, between the 13th and 16th centuries, met the needs of the small Italian city-states to focus on doing business, at the cost of a form of warfare in which a condottiero could change allegiance at a moment’s notice depending on who offered the most, much as a football manager or player might do today; and, moreover, it also happened that, to minimise losses, the mercenary companies hired by opposing sides on the battlefield would meet to reach an agreement behind closed doors, assessing who would win based on an evaluation of their respective potentials. As a response to feudal chivalry, which was tied to a world in decline, and to municipal militias that were too costly, the mercenary companies were in turn rendered obsolete with the creation of modern states under absolute monarchies. But by then, Western colonial expansion had been entrusted to a diverse array of conquistadors, privateers and East India companies, some of which lasted well into the late nineteenth century.

Apart from that, Severini observes that “at first glance, this trend seems human” – “fewer soldiers, fewer deaths”. But this risks altering the political costs of war, making them more acceptable, even routine. Thus, “endless wars arise from this discrepancy between military effort and political consequences”.