How Is AI Changing Modern Warfare?
AI is changing modern warfare not primarily through killer robots or sci-fi weapons — but through the speed and quality of decision-making. Two armies with identical weapons and hardware can produce radically different outcomes on the battlefield if one is powered by artificial intelligence and the other is not. The AI-enabled side can gather more intelligence, process it faster, identify targets more accurately, and execute strikes in a fraction of the time. That gap is the new definition of military superiority. Vladimir Putin understood this early. In 2017, he told a group of Russian students: "Whoever leads in AI will rule the world." The White House recently echoed that sentiment, issuing its first memo on AI and calling it an era-defining technology with significant and growing relevance to national security.
To understand how this actually plays out, consider a hypothetical — but deeply realistic — scenario: two armies with identical weapons face off. One has AI integration across its command and sensor network. The other does not. What happens next reveals everything about where warfare is heading.
01:15
The hypothetical battlefield setup: two identical armies, one with AI, one without
Watch at 01:15 →
What Is the Military Kill Chain?
Before understanding AI's role, you need to understand the kill chain — the sequence of steps every military must execute to successfully strike a target. It goes like this: first, you get eyes into enemy territory using sensors. Then you find the most valuable targets, fix their precise location, decide which weapons to use, deploy and fire those weapons, and finally assess whether the strike worked. That's the kill chain. And executing it is anything but simple.
Commanders on both sides are constantly making decisions with incomplete, unreliable, or overwhelming amounts of information. Victory consistently favors whoever can gather the best intelligence and make the most accurate decisions the fastest. That's exactly where AI enters — not at the point of pulling the trigger, but at every earlier step in the chain where data gets collected, analyzed, and acted upon.
The AI Advantage Starts With Computer Vision
Both armies in our hypothetical have sensors everywhere — satellites with cameras, drones scanning the terrain, radar tracking aircraft, reconnaissance teams on the ground. Both are generating enormous streams of visual data. But here's where the divergence begins. The AI-enabled command center runs all of that incoming imagery through computer vision software — the same underlying technology that learned to recognize cats in photos back in 2012. If it can recognize a cat, it can recognize a tank. The system automatically tags objects, assigns probability scores, and logs precise coordinates — in real time, across every sensor feed simultaneously. The non-AI army? They have analysts manually poring over every satellite image. Slowly. A tank here, an airfield there, plotted by hand onto a map that only exists in one command center.
04:30
How AI fuses sensor data — satellites, drones, phone records — into a single intelligence picture
Watch at 04:30 →
Meanwhile, the AI army's findings are being uploaded automatically to a shared cloud database accessible across all command centers. They're also fusing non-visual data — mobile phone records, ship tracking websites, aircraft transponders, even retail purchasing patterns near military bases that might signal unusual activity. All of it gets synthesized into a single intelligence picture, updated continuously.
How Do Autonomous Drones Work in Combat?
Drones are perhaps the most visible example of AI in modern combat — and the most misunderstood. The real breakthrough isn't that drones can fly. It's that they can now complete their mission without a human pilot when enemy jamming cuts the signal. Traditional drones are piloted remotely, but enemy forces can flood the airwaves with radio interference, severing the connection between pilot and drone mid-flight. That used to mean a lost drone and a failed mission. Now, AI changes that equation entirely.
07:10
Autonomous drone locking onto a target after signal jamming cuts the pilot's connection
Watch at 07:10 →
Thanks to advances in on-board computer vision, a drone can lock onto its target in the final 100 to 300 meters of flight and guide itself in — no pilot required. It compares what it sees with a library of images stored on the device itself, identifies the target, and maneuvers autonomously to strike it. This isn't science fiction. It's happening right now, and it represents what defense analysts call a genuine shift in the character of warfare — precision-guided autonomy available not just on expensive military hardware, but on cheap, mass-produced systems across the entire battlefield.
How Is AI Being Used in Ukraine and Gaza?
Ukraine has become the world's most important real-world test bed for military AI. Both sides are experimenting with and deploying AI-enabled systems at scale. The most significant battlefield innovation has been in drones — specifically, software that allows them to complete their final attack run even after signal jamming has cut the pilot's connection. Computer vision running locally on the drone finishes the job.
10:45
Real-world examples: AI targeting systems in Gaza and drone AI in Ukraine
Watch at 10:45 →
In Gaza, the Israeli military has deployed an AI targeting system capable of generating large numbers of potential targets per day — identifying suspected militants based on signals intelligence, location data, and pattern-of-life analysis. According to reporting based on accounts from six Israeli soldiers, the IDF has relied heavily on AI-generated target lists, with operators sometimes treating the machine's output as if it were a human decision. The IDF disputes that characterization, but the underlying reality is undeniable: AI algorithms are already being used to choose targets in active combat zones.
What Is the US Army's Project Convergence?
The US military's answer to all of this is an ambitious initiative called Project Convergence — an attempt to aggregate data from every sensor across the battlefield into a single integrated software platform that helps commanders make faster, better-informed decisions. The goal is a unified command interface where satellite imagery, drone feeds, signals intelligence, logistics data, and target recommendations all flow into one place. The Army describes it as more efficient and faster than humans operating the same functions manually. Data analytics firm Palantir — which works extensively with the US military — claims their software allows operations that once required 2,000 personnel to be run by just 20, because the software handles the analysis and synthesis that used to require entire teams.
The UK has a similar aerial imagery program, and cloud computing giants like Amazon and Microsoft are increasingly involved in defense infrastructure, providing the processing power needed to run these systems at scale.
Is Using AI in Warfare Ethical or Safe?
This is the part that should keep everyone up at night — and it's more complicated than the usual "killer robots" framing suggests. The most common argument is that a human must always retain control over the final decision to use lethal force. But Economist Defense Editor Shashank Joshi makes a sharper point: the really consequential decision isn't pulling the trigger. It's identifying something as a target in the first place. If an AI system generates a target list and a commander just rubber-stamps every recommendation without genuine intellectual or moral engagement, that's not meaningful human control — even if technically a human pressed the button.
What actually matters, Joshi argues, is whether the humans in the loop understand the system they're trusting. Do they know what it was trained on? Do they understand its blind spots? Do they know when to question it, and when to override it? A commander who deeply understands the AI's capabilities and limitations — and critically evaluates its outputs — represents genuine human control. A commander mindlessly approving AI-generated kill lists at scale does not, regardless of how you label it procedurally.
There's also the hallucination problem. AI systems make mistakes. They will, at some point, recommend striking a building believed to be a military target that turns out to be a civilian shelter. Humans make that mistake too. The real question is whether AI reduces the frequency of those catastrophic errors — or whether the speed and authority of machine-generated outputs makes commanders less likely to question them when they're wrong.
Why Did Putin Say AI Will Rule the World?
Putin's statement to Russian students in 2017 wasn't hyperbole — it was a strategic assessment. Military power has always been tied to technological advantage. Gunpowder, the industrial revolution, nuclear weapons — each reshuffled the global order. AI is the next reshuffling. The nation that can most effectively integrate AI into its military decision-making will be able to execute the kill chain faster, with greater precision, at lower cost, and at a scale no human-only force can match.
That's why the US issued its first national security memo specifically on AI. That's why tech startups focused on defense AI are attracting massive investment. And that's why no one — no government, no military — is going to voluntarily slow down this development and wait for ethicists and lawyers to settle the questions. Geopolitics doesn't work that way. The race is already on. The only question left is whether the humans who control these systems will be wise enough to understand what they've built — and humble enough to know when not to trust it.
