A question appears more and more often: Can artificial intelligence start a war? The question sounds dramatic, but it reveals a misunderstanding. Artificial intelligence does not want anything. It has no fear, ideology or national interest. Wars are not started by technology. They are started by people. Gunpowder didn’t start wars. Fighter jets didn’t start wars. Nuclear weapons didn’t start wars. – People did.
The rise of AI defence technology
What is changing is the role of AI in modern warfare.
A new industry — defence tech — has emerged. Companies such as Anduril Industries, Palantir Technologies, Shield AI and Skydio are building systems that analyze data, process satellite imagery, control drones and provide battlefield intelligence.
In effect, they are creating the operating system of modern warfare.
Speed — the real change
As I mentioned, AI does not decide to start wars. But it changes how fast decisions are made.
Satellite images, sensor data and troop movements can now be analysed in seconds. That speed can improve awareness — but it can also reduce time for reflection.
And wars have often begun in moments when reflection disappears.
Power, profit and technology
Technology alone does not start wars. But it can make them easier to justify and easier to sell.
The global defence industry is worth trillions of dollars. Where that much money flows, idealism rarely leads the conversation.
AI has simply become the newest frontier in that race. Winning the technology race does not automatically make the world safer. Sometimes it simply means we have invented faster ways to make catastrophic mistakes.
Responsibility still belongs to humans
The real risk is not that machines become autonomous.
The real risk is that human responsibility fades behind algorithms.
If an AI-assisted system makes a fatal mistake, who is responsible?
- The engineer?
- The company?
- The commander?
- The politician?
AI can analyse the world faster than ever. But it does not carry moral responsibility.
That still belongs to us.
