Let me tell you something that gives me chills: imagine a machine deciding whether to bomb a school based on a spreadsheet. That’s not sci-fi—it’s happening now. The US military’s Maven Smart System, which helped strike over 1,000 targets in the first days of the Iran war, including a school in Minab, is a case study in the absurdity of trusting algorithms with human lives. Personally, I think this isn’t just a technical problem—it’s a moral catastrophe waiting to happen. What makes this particularly fascinating is how we’ve normalised the idea that speed and efficiency trump empathy. If you take a step back and think about it, we’re essentially outsourcing our conscience to code. This raises a deeper question: Are we building weapons or erasing the very humanity that should guide warfare?
The rise of AI in militaries isn’t just about drones or targeting systems—it’s about rewriting the rules of engagement. Global spending on military AI is projected to jump from $11.7bn to $19.3bn by 2030. That’s not just numbers; it’s a cultural shift. What many people don’t realize is that this isn’t just about technology—it’s about power. The Pentagon’s push for an ‘AI-first fighting force’ isn’t about winning wars faster; it’s about consolidating control. A detail that I find especially interesting is how they’re reviving the OODA loop, a 20th-century concept that prioritizes speed over deliberation. In my opinion, this is a dangerous game of chess where the pieces are lives, and the board is a battlefield.
Now, let’s talk about Maven. This system isn’t just processing data—it’s weaponizing it. It pulls from satellites, social media, and even radio emissions to create a kill chain that moves faster than a human could ever react. But here’s the kicker: the error rate isn’t disclosed. The Minab strike, which killed over 150 people, was traced back to outdated satellite intel. What this really suggests is that we’re relying on systems that can’t even account for their own mistakes. And yet, we’re told this is the future. From my perspective, this is like trusting a drunk driver with a GPS. The system might be precise, but the inputs are flawed.
Elke Schwarz, a political theorist, argues that AI can’t grasp morality because it lacks the capacity to feel human suffering. She’s right, but it’s more than that. The real problem is how we’ve trained ourselves to accept this. When the Pentagon slashes its civilian protection team from 40 to 9 staff while relying on AI for harm assessments, it’s not just cutting costs—it’s abdicating responsibility. A hidden implication here is that we’re creating a feedback loop where AI’s ‘efficiency’ justifies fewer safeguards. If you think about it, this is the same mindset that allowed the drone wars of the 2000s to escalate: ‘We’re just following the algorithm.’
And then there’s the paradox of AI as a moral check on unethical militaries. Some argue that AI could constrain rogue states by enforcing rules of engagement. But Schwarz isn’t buying it. She’s spot-on when she says an AI won’t stop a military from doing what it wants—it’ll just make it faster. The Israeli use of Lavender, an AI system with a 10% error rate, to target Palestinians is a chilling example. What this shows is that AI doesn’t fix the problem; it amplifies it. The real danger isn’t the machines—it’s the humans who think they can outsource morality to software.
So where do we go from here? I’m not sure, but one thing is clear: the line between tool and tyrant is getting blurrier by the day. We need to ask ourselves if we’re building a future where machines decide who lives and dies, or if we’re finally learning that some decisions—like whether to bomb a school—should never be left to code. The answer will define not just our wars, but our humanity.
