2nd October 2026
The Pentagon has just announced something that could prove more important than another new weapons system.
America is creating an Autonomous Warfare Command, or AUTOWARCOM, a new four-star military command designed to accelerate the development and deployment of drones, artificial intelligence and robotic weapons across the US armed forces.
The command is expected to be established by October 2027 and will have what the Pentagon describes as “service-like authorities”. The intention is to overcome one of the traditional problems of military procurement: by the time an expensive new weapons system has been designed, tested and introduced, the technology may already be moving on.
The Pentagon's argument is straightforward.
Modern warfare is changing rapidly, particularly because of the experience of Ukraine. Cheap drones can perform reconnaissance, carry explosives, attack targets and operate in large numbers.
The traditional military model has often been based on a relatively small number of extremely expensive platforms.
AI and autonomous systems offer something different.
Mass.
Instead of asking how many £50 million aircraft or £5 million missiles a country can afford, military planners can increasingly ask how many thousands of relatively cheap autonomous systems can be produced and deployed.
That could change the economics of warfare.
Britain is already moving in the same direction
The UK does not have to wait for America to demonstrate the idea.
Britain's own 2025 Strategic Defence Review explicitly says that modern warfare is increasingly based on networks combining crewed, uncrewed and autonomous systems. It calls for a major expansion of AI and autonomous capabilities and says these technologies can provide greater accuracy, lethality and cheaper capabilities.
The review recommends incorporating uncrewed and autonomous systems into the Armed Forces in high numbers over the following five years.
It even proposes a possible 20-40-40 mix for land warfare: 20% crewed platforms, 40% reusable systems and 40% expendable systems such as drones and other one-way weapons.
And this is no longer simply a plan on paper.
In September, Britain's Defence Science and Technology Laboratory announced that the Army had successfully experimented with an autonomous drone swarm consisting of eight uncrewed aircraft.
The UK is also working with Ukraine to use real battlefield data to develop AI models for future drone swarms.
So Britain is already heading towards what America is now calling “Autowar”.
But there is a much bigger question
If autonomous weapons become significantly cheaper than traditional weapons, what happens to the cost of going to war?
That could have two completely different effects.
On one hand, cheaper systems could make defence more affordable. Britain could potentially put thousands of drones into the air for the price of a handful of conventional aircraft.
But the other possibility is much more uncomfortable.
If war becomes cheaper, could it become easier to continue?
A country that has lost aircraft and highly trained personnel faces a significant cost when replacing them.
Losing an inexpensive drone is different.
An autonomous system could potentially be replaced within days or weeks, particularly if it is manufactured using commercial technology.
That could encourage a shift from protecting a limited number of extremely expensive platforms towards deploying enormous numbers of expendable ones.
Ukraine has already demonstrated elements of this transformation.
Other countries will have to respond
America's decision creates pressure on its allies.
If the US can deploy large numbers of autonomous systems, Britain, France, Germany and other NATO members will have to consider whether their forces could operate effectively alongside them.
There is also the opposite problem.
What happens if a potential adversary develops thousands of autonomous drones while Britain and its allies are still relying heavily on conventional platforms?
That is why the UK's Strategic Defence Review places such emphasis on autonomy, AI, counter-drone systems and the ability to manufacture munitions and equipment rapidly.
The result could be a new military arms race.
But it may not look like the Cold War arms race.
Instead of competing over who has the most nuclear missiles or aircraft carriers, countries could increasingly compete over who can design, manufacture and deploy the largest numbers of intelligent machines.
And that brings another problem.
The machines themselves need to be defended.
An army deploying thousands of drones will also need systems capable of detecting and destroying an enemy's drones. Electronic warfare, cyber warfare, sensors and directed-energy weapons therefore become increasingly important.
Britain's DragonFire laser programme is one example of the search for cheaper ways of defeating relatively cheap incoming drones.
And then there is the human question
The most difficult issue may not be technological at all.
It is deciding how much authority should be given to machines.
The UK Government says its autonomous weapons will comply with international humanitarian law and that human involvement remains important.
But as systems become faster and more autonomous, the boundary between a human making a decision and a machine making one becomes increasingly important.
If an autonomous system identifies a target incorrectly, who is responsible?
And if two countries deploy autonomous systems that react to each other's actions at machine speed, how much opportunity is there for humans to stop an escalation?
These questions become particularly serious when AI systems are operating across land, sea, air, space and cyberspace simultaneously.
The cost of war may be changing
America's creation of AUTOWARCOM therefore deserves attention far beyond the Pentagon.
It represents an attempt to reorganise a major military around a new technological reality.
Britain is already moving in the same direction.
Other countries will almost certainly have to consider how they respond.
The extraordinary possibility is that the next generation of warfare could make some military capabilities dramatically cheaper.
But that does not necessarily mean a cheaper world.
If the cost of deploying military force falls, the threshold for using it could fall too.
The question for governments may therefore become not simply “Can we afford this war?”
It could increasingly be:
“If machines make war cheaper, will that make countries more willing to fight?”
That may ultimately be the most important question raised by America's new Autowar command.
So once again, mankind seems willing to utilise the amazing possibilities of AI for lethal purposes, rather than simply using it to help ensure that people can live better and more fulfilling lives.