ASIA
30 August 2026
AI, Automation and War
Will AI change the nature of war?
Will Artificial Intelligence (AI) automate the theatre of war and displace political leaders and human commanders? Will the future battlespace be dominated by drone swarms and killer robots, controlled by sophisticated algorithms and programmed with massive amounts of data?
Will war become a competition between AI and lethal autonomous systems, rather than between people? In other words, will AI transform warfare just as gunpowder or airpower did? It’s not so simple, according to Anthony King in his excellent book, “AI, Automation, and War: The Rise of a Military-Tech Complex”.
King worries that many proponents of AI seem to exaggerate its capabilities. In particular, AI leaders from Silicon Valley make extraordinary hyperbolic claims about the powers and capabilities of AI, and simultaneously about the extreme dangers that this poses. A lot of these claims are by interested parties, looking for protectionist regulation from the government. In reality, AI merely operates on statistical probabilities and pattern recognition. It lacks real political judgment, intuition, and the creative adaptation required to navigate the "fog and friction" of combat.
All the leading military powers have recognised the potential of AI over the past 10-20 years. But autonomous drone swarms and robots have not been the focus of these efforts by the armed forces. AI has been used to mine the vast resources of cyberspace, and has been employed for three principal military functions: planning, targeting, and cyber operations. In this regard, AI requires greater human teamwork, not less.
Ukrainian forces were able to use data from satellites, open sources, and encrypted signals intelligence to plot and attack the precise location of Russian command posts and logistics hubs hundreds of miles behind the front line, killing around 30 Russian officers and destroying large numbers of logistic hubs.
AI models are vital to all kinds of cyber operations, as they can help the armed forces to hack into an enemy’s systems—to carry out cyber sabotage, espionage, and subversion—and to protect their own systems from penetration. AI has also been increasingly employed for information and psychological operations.
Thus, over the last decade, ministries of defence have reformed the regulatory environment to facilitate cooperation between the tech and defence sectors. Defence ministries have signed contracts with Microsoft, Google, Amazon Web Services, and other tech companies to provide them the services they require. A new ecology is appearing: a military-tech complex, which is altering – and, in some cases, displacing—the old military-industrial complex.
Unlike the arms companies of the twentieth century, however, tech companies have not developed software platforms independently of the armed forces and then simply sold these products to them. In order to apply AI to specific operational problems, data engineers and programmers require immediate access to the available data.
The armed forces have thus integrated data scientists and programmers into operational headquarters, where they can work together with commanders and staff officers. AI is just a very sophisticated tool for these human teams, one which is programmed, trained, and refined to process specific bodies of identified data in order to answer particular questions set by human commanders.
Elon Musk’s Starlink communications system played perhaps the most vital role. At the beginning of the war, Ukraine appealed personally to Musk for assistance. Within days of the invasion, the company was providing communications for Ukrainian special operations brigades and connecting them with the US Joint Special Operations Command.
Starlink was crucial to Ukraine’s war effort. It allowed the government and the armed forces to communicate securely. The Ukrainians fully admitted their dependence on Musk: ‘Without Starlink, we would have been losing the war’, a Ukrainian officer said.
In September 2022, Musk temporarily withdrew Starlink services from Ukraine. He was concerned that the Ukrainians were planning to launch a surprise attack against Russian ships in Sevastopol, using uncrewed surface vessels packed with explosives. He believed that such a move was potentially escalatory.
The fact that the operation was impossible without Starlink demonstrated just how important Musk’s company was to the Ukrainian war effort. It seems very likely that Microsoft, Google, Amazon Web Services, and Palantir have probably all influenced US policy on Ukraine in some way.
The rise of a military-tech complex amounts to a quasi-privatisation of national defence strategy itself. Tech primes are becoming part of the strategic community, they are becoming political actors. Indeed, it is deeply ironic: AI has not automated strategy, as many have feared, but politicised it. Private tech companies are now involved in strategy too.
Since the Israel Defence Force’s withdrawal from Gaza in 2005, the Israelis constructed one of the most sophisticated and intrusive spying systems, the purpose of which was to monitor Palestinian, especially Hamas, activity. By 2021, they had established an elaborate architecture of digital surveillance.
AI played an important role in collating and processing the data flowing in from open-source digital sites, from a forest of cameras, from satellites, and from signals intelligence. Yet, on 7 October 2023, the Israeli digital surveillance system, and all its AI agents and software, failed. They remained mute before and during the attack. This was because Hamas had employed about forty home-made suicide drones to destroy a network of remotely controlled weapons, sentry towers, cameras, and communications devices.
In short, AI is as limited and vulnerable as any other form of military technology. The mistakes of AI have often been so egregious that it is obvious when AI generates a false result. Yet, in some cases, AI produces answers that are mistaken but plausible. Commanders have to be aware that, just like humans, AI is fallible.
In the next decade, warfare is not going to fulfill the fantasies of science-fiction writers, writes King. Cyborgs will not take over. Supercomputers and killer robots and drones will not replace humans. Nevertheless, it is imperative that no one be under any illusion. The development of AI and its accompanying military-tech complex will change the armed forces, the way they fight, and their relations to the state and society.
Will war become a competition between AI and lethal autonomous systems, rather than between people? In other words, will AI transform warfare just as gunpowder or airpower did? It’s not so simple, according to Anthony King in his excellent book, “AI, Automation, and War: The Rise of a Military-Tech Complex”.
Getting realistic about AI
King, a professor of war studies and director of the Strategy and Security Institute at the University of Exeter, brings a deeply nuanced and empirical perspective to these questions. He argues that strategists and military scientists have regularly exaggerated the revolutionary potential of new technologies like the strategic bomber. For example, strategic bombing in the Second World War was important, but not strategically decisive.King worries that many proponents of AI seem to exaggerate its capabilities. In particular, AI leaders from Silicon Valley make extraordinary hyperbolic claims about the powers and capabilities of AI, and simultaneously about the extreme dangers that this poses. A lot of these claims are by interested parties, looking for protectionist regulation from the government. In reality, AI merely operates on statistical probabilities and pattern recognition. It lacks real political judgment, intuition, and the creative adaptation required to navigate the "fog and friction" of combat.
All the leading military powers have recognised the potential of AI over the past 10-20 years. But autonomous drone swarms and robots have not been the focus of these efforts by the armed forces. AI has been used to mine the vast resources of cyberspace, and has been employed for three principal military functions: planning, targeting, and cyber operations. In this regard, AI requires greater human teamwork, not less.
AI in practice
AI may be an advanced technology, but it is better understood as a complex suite of capabilities. It requires data, computing power, and expertise. King recounts an exquisite case of the use of AI in the Russo-Ukraine War, thanks to direct assistance from the US. This was “Task Force Dragon” directed from a secret command center in Germany.Ukrainian forces were able to use data from satellites, open sources, and encrypted signals intelligence to plot and attack the precise location of Russian command posts and logistics hubs hundreds of miles behind the front line, killing around 30 Russian officers and destroying large numbers of logistic hubs.
AI models are vital to all kinds of cyber operations, as they can help the armed forces to hack into an enemy’s systems—to carry out cyber sabotage, espionage, and subversion—and to protect their own systems from penetration. AI has also been increasingly employed for information and psychological operations.
Partnerships with Silicon Valley
In order to leverage AI, the armed forces have had to forge an alliance with Silicon Valley, according to King. Only the tech sector has the data, the expertise, the software, the computing power, and the capital, provided by venture capitalists, to help the armed forces develop and exploit AI.Thus, over the last decade, ministries of defence have reformed the regulatory environment to facilitate cooperation between the tech and defence sectors. Defence ministries have signed contracts with Microsoft, Google, Amazon Web Services, and other tech companies to provide them the services they require. A new ecology is appearing: a military-tech complex, which is altering – and, in some cases, displacing—the old military-industrial complex.
Unlike the arms companies of the twentieth century, however, tech companies have not developed software platforms independently of the armed forces and then simply sold these products to them. In order to apply AI to specific operational problems, data engineers and programmers require immediate access to the available data.
The armed forces have thus integrated data scientists and programmers into operational headquarters, where they can work together with commanders and staff officers. AI is just a very sophisticated tool for these human teams, one which is programmed, trained, and refined to process specific bodies of identified data in order to answer particular questions set by human commanders.
Politicisation of military-tech complex
Because tech companies are involved in operations themselves, they exert an influence which the traditional defence industries, building only platforms, rarely did, writes King. The Russo-Ukraine War has exemplified the transformation. In February 2022, in the immediate aftermath of the invasion, many tech primes, including Google, Microsoft, and Amazon Web Services, supported the Ukrainian government.Elon Musk’s Starlink communications system played perhaps the most vital role. At the beginning of the war, Ukraine appealed personally to Musk for assistance. Within days of the invasion, the company was providing communications for Ukrainian special operations brigades and connecting them with the US Joint Special Operations Command.
Starlink was crucial to Ukraine’s war effort. It allowed the government and the armed forces to communicate securely. The Ukrainians fully admitted their dependence on Musk: ‘Without Starlink, we would have been losing the war’, a Ukrainian officer said.
In September 2022, Musk temporarily withdrew Starlink services from Ukraine. He was concerned that the Ukrainians were planning to launch a surprise attack against Russian ships in Sevastopol, using uncrewed surface vessels packed with explosives. He believed that such a move was potentially escalatory.
The fact that the operation was impossible without Starlink demonstrated just how important Musk’s company was to the Ukrainian war effort. It seems very likely that Microsoft, Google, Amazon Web Services, and Palantir have probably all influenced US policy on Ukraine in some way.
The rise of a military-tech complex amounts to a quasi-privatisation of national defence strategy itself. Tech primes are becoming part of the strategic community, they are becoming political actors. Indeed, it is deeply ironic: AI has not automated strategy, as many have feared, but politicised it. Private tech companies are now involved in strategy too.
AI’s shortcomings
AI is not fail-safe, King insists. It is as prone to error and omission as any other source of military intelligence and information. An egregious case of AI failure occurred on 7 October 2023 when thousands of Hamas fighters streamed out from Gaza into southern Israel, under the cover of a massive rocket barrage.Since the Israel Defence Force’s withdrawal from Gaza in 2005, the Israelis constructed one of the most sophisticated and intrusive spying systems, the purpose of which was to monitor Palestinian, especially Hamas, activity. By 2021, they had established an elaborate architecture of digital surveillance.
AI played an important role in collating and processing the data flowing in from open-source digital sites, from a forest of cameras, from satellites, and from signals intelligence. Yet, on 7 October 2023, the Israeli digital surveillance system, and all its AI agents and software, failed. They remained mute before and during the attack. This was because Hamas had employed about forty home-made suicide drones to destroy a network of remotely controlled weapons, sentry towers, cameras, and communications devices.
In short, AI is as limited and vulnerable as any other form of military technology. The mistakes of AI have often been so egregious that it is obvious when AI generates a false result. Yet, in some cases, AI produces answers that are mistaken but plausible. Commanders have to be aware that, just like humans, AI is fallible.
Conclusion
The emerging military-tech complex transforms the way in which states defend themselves and fight each other. The increasing reliance of the armed forces on the data, the expertise, the computing power, and the capital of the tech sector has profound implications not just for the armed forces but for civil-military relations themselves. It revises the political settlements which were established between governments, industry, and the armed forces in the twentieth century.In the next decade, warfare is not going to fulfill the fantasies of science-fiction writers, writes King. Cyborgs will not take over. Supercomputers and killer robots and drones will not replace humans. Nevertheless, it is imperative that no one be under any illusion. The development of AI and its accompanying military-tech complex will change the armed forces, the way they fight, and their relations to the state and society.