Die ferngesteuerten Drohnen und die computergestützte Zielerfassung hinter den Schrecken des modernen Krieges

As the world nervously awaits the next move in the ongoing war between [Israel](https://inews.co.uk/topic/israel?ico=in-line_link), the US and [Iran](https://inews.co.uk/topic/iran?ico=in-line_link), fast-evolving [artificial intelligence](https://inews.co.uk/topic/ai?ico=in-line_link) systems are helping to drive almost every serious military decision.

These processes that have gradually evolved since [9/11](https://inews.co.uk/topic/9-11-attacks?ico=in-line_link), helping to track down the likes of Saddam Hussein and [Osama bin Laden](https://inews.co.uk/topic/osama-bin-laden?ico=in-line_link) and later finding, monitoring and taking out targets in conflicts around the world.

At its heart is a simple reality: modern AI-enabled systems allow more individuals, groups, devices and systems to be identified and targeted, but interpreting the data is often fraught with dangerous challenges.

Firms that handle this dynamic – including growing powerhouses like Palantir – have become extremely influential, but even they now struggle to control the often-messy dynamics this new technology produces.

For all the talk by military commanders and the tech firms they hire of keeping “humans in the loop”, the speed with which AI-enabled targeting can allow decisions to be taken is already driving wider strategy.

Sometimes – as in the current conflict – that can deliver much greater precision: the death toll related to the Iran war so far is significantly lower than a region-wide war might have inflicted in previous decades.

But when the targets are individuals living in teeming cities like Gaza, Tehran or Beirut, mistakes can have immediate and appalling human consequences. This is already rumoured as one of the reasons for a [US Tomahawk missile strike on a school for girls in Iran](https://www.washingtonpost.com/national-security/2026/03/11/us-strike-iran-elementary-school-ai-target-list/) late last month, which is said to have killed at least 175 people, mostly students.

Getting to the truth of individual incidents is often incredibly hard. That includes what one UN probe described as perhaps the first-ever example of attack drones automatically targeting vehicles and possibly even soldiers fleeing a town in Libya in 2020, when the world’s attention was distracted by the Covid-19 pandemic.

It may now be too late to ever know for sure what happened in that incident. What is clear, however, is that the growing computing power, international tensions and array of defence technology companies are changing the face of war.

In the Pacific, US commanders have [talked of hitting “a thousand” high-value targets](https://www.congress.gov/crs-product/IF12611) in the first 24 hours of any conflict. Once “kinetic action” – the US euphemism for military activity – starts, however, it becomes inevitable that any target lists will include mistakes, what [some refer to as “hallucinations”](https://www.afcea.org/signal-media/18th-airborne-corps-tricky-ai-journey), where AI algorithms look at the data and come to a completely wrong conclusion.

They can also deliver complacency. In late 2023, barely a week before [Hamas](https://inews.co.uk/topic/hamas?ico=in-line_link) launched its deadly attack on Israel on 7 October, Israeli forces invited their Nato counterparts to inspect military and surveillance facilities, fences, automated weapons systems and high-tech sensors that separated the [Gaza Strip](https://inews.co.uk/topic/gaza?ico=in-line_link) from the outside world.

The high-tech solutions had lured Israeli commanders into a false sense of security. As frontline surveillance [reported Hamas fighters visibly preparing for an assault](https://www.bbc.co.uk/news/world-middle-east-67958260) – even roleplaying attacks on Israeli tanks – military chiefs continued to wrongly put their trust in their computer-based algorithms and analysis.

The results would prove devastating, first for Israeli communities and those close to Gaza when the assault was launched, and then for Gaza over the years-long Israeli military campaign that followed.

Israeli military sources have also told Israeli media that they believe AI programs likely massively increased the death toll in Gaza – through the side effects of having technology deliver large numbers of potential targets.

Since 2023, AI and machine learning programs, along with data aggregation, analysis and targeting software, have allowed pretty much every individual and mobile phone in Gaza to be profiled almost instantaneously against the likelihood they are linked to Hamas. Even if human analysts decided to remove one or more targets from the list because they doubted their veracity, there were thousands more remaining.

Similar tactics appear to have been used in Lebanon to target [Hezbollah](https://inews.co.uk/topic/hezbollah?ico=in-line_link) sites. And much more recently, both Israel and the US appear to have used these systems to detect targets within Iran itself, identifying top commanders and decision-makers. Israel has launched strikes against many of Iran’s top leaders.

Going forward, there will be very little choice for nations to turn their back on these technologies, nor the increasingly overlapping constellations of firms and international alliances that use them every day. Just like barbed wire, machine guns, artillery, mortars and even the simple handgun and bayonet – all still used in the current century – AI targeting and remote drones are simply facts of modern warfare.