This budget request is either a long-overdue wake-up call, or the most expensive way imaginable to admit we didn’t take drones seriously soon enough.
From what’s been shared publicly, the Pentagon is asking for a massive surge in funding for its Defense Autonomous Warfare Group—$54.6 billion for fiscal year 2027, up from $225 million the prior year. That’s not a normal increase. That’s a sprint. And it’s being justified by a blunt lesson coming out of recent wars: low-cost “kamikaze” drones work. They’re cheap, they’re hard to stop, and they can change the outcome of a fight faster than a lot of traditional hardware.
As a company that builds drone detection radar systems and AI fusion from different sensors, we should be cheering. More funding usually means more deployments, more demand, and more urgency. But I’m not going to pretend this is automatically good news. A budget spike like this can mean serious progress—or it can mean panic spending that buys the wrong things at the wrong scale, then leaves everyone wondering why the threat still got through.
Here’s the uncomfortable part: the problem isn’t that drones exist. It’s that they don’t behave like the threats most bases, convoys, and ships were built to handle. A small, fast, low-flying drone can slip into the gaps—gaps in coverage, gaps in rules, and gaps between teams who each “own” a slice of the problem. One team has a sensor. Another team has a jammer. Another team has a weapon. Another team has the authority to decide what to do. The drone doesn’t care about your org chart.
So when I see $54.6 billion aimed at “autonomous warfare,” I read it as the Pentagon trying to buy speed. Not just more drones, but the ability to find them, track them, decide fast, and respond. That’s the right direction. But spending big doesn’t guarantee you get there.
The first thing that can go wrong is simple: people buy shiny offense and treat defense like an accessory. Offense has the sexy demos. Defense has the boring work—coverage maps, false alarms, training, maintenance, and the reality that the environment is messy. Radar drone detection in the real world isn’t a clean lab test. You have birds. You have terrain. You have weather. You have clutter. If leadership expects “one sensor” to solve it, they’ll waste money and lose time.
The second failure mode is even more dangerous: building a system that sees everything but can’t act. Detection without decision is just anxiety with a dashboard. Imagine a forward base where the screens light up with “possible drone” alerts all night. The operators start ignoring it because half the time it’s nothing. Then the one time it matters, the response is late. That’s not a technology failure. That’s a design failure—and it happens when people treat sensor data as the product, instead of the outcome: fewer successful attacks.
This is where our perspective matters. In our world, the win is not “more sensors.” The win is making sensors work together so the system is confident enough to act, and calm enough to be trusted. AI fusion from different sensors is not a buzz phrase to us—it’s the difference between a radar hit that might be a bird and a radar hit that lines up with other signals and becomes a real track that someone can respond to quickly.
Still, I’m wary of what a 243x increase does to incentives. When budgets explode, timelines get heroic. Requirements get vague. Everyone claims their product is “autonomous” because that’s where the money is. And the people writing checks often don’t have the patience for the hard truth: defense against cheap drones can’t be solved with one expensive miracle box. It’s a grind. It’s layers. It’s constant updates. And it’s accepting that the attacker gets a vote.
There’s also a human cost to getting this wrong. Picture a ship captain deciding whether a blip is a drone or a civilian object. Picture a base commander choosing whether to jam an area and accidentally knock out friendly systems. Picture a convoy that has to move anyway because the mission doesn’t pause, even when drones are in the area. If the detection system is noisy, you either freeze—and lose operational freedom—or you push through and accept hits. Neither is a “good” option.
On the flip side, if this money is spent with discipline, it could actually make conflict less deadly. If low-cost drones are now the weapon everyone can afford, then affordable defense becomes a kind of stability tool. It keeps small attacks from turning into big escalations. It gives commanders options besides “shoot first” or “do nothing.” That’s a big deal.
But I can’t ignore the bigger question underneath all of this: are we funding autonomy because it’s the best path, or because it’s the only way to keep up with how fast the battlefield is changing? Those are not the same motivation. One leads to careful building. The other leads to rushing, overpromising, and fielding systems that look good in a briefing and fail when conditions get ugly.
If the Pentagon really wants results, the measure shouldn’t be how many autonomous systems it buys. It should be whether drone attacks stop being “surprising,” whether response gets faster without becoming reckless, and whether units trust what the system tells them when it matters.
So here’s the debate I want people to have out loud: if you had to choose, would you spend the next wave of this money on more autonomous attack systems, or on scaling reliable defense—radar drone detection and sensor fusion that actually works under pressure—so the battlefield stops rewarding the cheapest drone with the luckiest flight path?