The era of high-tech assassinations, conducted almost entirely remotely with terrifying precision, may be upon us.

This week, Ali Fadavi, the deputy commander of the Islamic Revolutionary Guard Corps, said a satellite-controlled machine-gun with "artificial intelligence" was used to murder a key Iranian nuclear scientist.

According to his account, Mohsen Fakhrizadeh was driving outside Tehran on November 27, when the machine gun "zoomed in" on his face and fired 13 rounds.

The technology enabled such precision that the scientist's wife escaped unharmed, even though she was only 10-inches away, he said.

Doubts persist over Fadavi's extraordinary claims. "There is no satellite that shoots a machine gun, nor are there satellites that leverage AI and remotely control a gun on the ground," says Missy Cummings, director of Duke University's Humans and Autonomy Laboratory.

"I think it's more likely that this was a remote-controlled weapon," says Ingvild Bode, professor at the University of Southern Denmark's Centre for War Studies.

"Maybe AI was used to support the targeting but I don't think that AI did the targeting itself, because the killing happened in a very kind of dynamic urban environment that makes it very unlikely for AI to be used precisely in this way."

Nevertheless, whether or not the exact details are correct, one fact remains: new military technologies are enabling more and more remote-controlled and autonomous weapons, making remote assassinations increasingly possible.

Bode says the use of technologies that enable remote assassinations are blurring the line between war and peace.

"Usually, if you're not living in a battlefield or in the state of war, you're supposed to be free from becoming the subject of violent force, but we've seen that drones have also been used outside of battlefields," she says.

Arthur Holland Michel, associate researcher at the United Nations Institute for Disarmament Research, cautions against believing unconfirmed technical details but says:  "The game has been changing for a very, very long time and [the assassination in Iran] may, in reality, be just one more step in that evolution."

Since the first known drone strike in Yemen in 2002, remote assassinations have been dominated by unmanned planes, including the assassination of Iranian general, Qasem Soleimani, in January by the United States.

The US Army has been experimenting with drone "swarms" which hunt in packs and can fan out to locate and destroy specified targets.

A competition document released in September showed the UK's Ministry of Defence was also searching for unmanned drones that could carry deadly weapons so British forces could “remove service personnel and military dogs from complex and dangerous urban warfare situations where their lives are put at significant risk”.

At the forefront of this march towards soldier-free warfare are weapons, equipped with artificial intelligence, that can decide on their own what - or who - to attack.

According to Paul Scharre in his book 'Army of None: Autonomous Weapons and the Future of War', more than 30 nations already have access to defensive autonomous weapons.

The Harpy kamikaze drone created by Israel Aerospace Industries, for example, is designed to detect and crash into the source of any enemy radar signals it detects, destroying both itself and the target without human intervention.

But drones aren't the only option being explored by nations.  The UK's chief of the defence staff recently suggested robot soldiers could soon make up a quarter of the army.

The government's Defence and Security Accelerator has also announced funding opportunities for companies that could offer security-focused "telexistence solutions" - a futuristic type of robotics that enables a person to teleport their movements to a robot based elsewhere.

In reality, such remote assassination tools operate at the messy boundaries of technical feasibility.

According to a senior Russian official speaking at a conference in 2018, when the country deployed it's Uran-9 remote-controlled tank in Syria, the technology repeatedly lost contact with the control station at one point for up to 1.5 hours and it operated best when just 300-500m away from the person controlling its movement.

Doubts over Iran's accounts of the satellite-controlled AI machine gun also suggest the technology isn't quite as advanced as it seems.

Although the US, Turkey and Israel already use remote-control guns to guard military outposts in remote areas, the weapons are usually overseen by human-beings stationed nearby, says Wim Zwijnenburg, an expert in drone and robot warfare at human rights organisation Pax for Peace.

Operators trying to communicate with drones via satellite over long distances can often encounter lags of between one or two seconds. Because drones usually fire missiles at static targets or large objects such as cars, "it's an explosion so it doesn't matter so much if it's like 10 or 15cm off", says Zwijnenburg.

The technology may still be evolving - and some claims need to be taken with a pinch of salt - but it's clear that nations are exploring these assassination technologies with greater zeal than ever before. The future of warfare could be remote, and there's no sign the tech arms race for increasingly sophisticated weaponry will slow down anytime soon.