AI Is Becoming the Military’s New Advisor, Whether We Like It or Not

MIT Technology Review just dropped an exclusive eBook bundle of six stories tracking how the world’s militaries are embedding AI into their decision loops. It’s not speculative future stuff—these pieces were originally published between April 2025 and April 2026, with updates for recent events. If you’ve been following defense tech, you know the shift has been accelerating faster than most civilians realize.

The collection covers terrain that’s uncomfortable but necessary to understand. One story examines how generative AI is learning to spy for the US military. Another features a defense official openly discussing how AI chatbots could be used for targeting decisions. That’s not a hypothetical. That’s someone with clearance telling us the machine is already in the room when lethal choices get made.

There’s also a piece on what the Pentagon is calling “phase two” of military AI—moving beyond simple data analysis into autonomous or semi-autonomous operational roles. And a separate story reveals the Pentagon is planning for AI companies to train their models on classified data. That raises obvious questions about security, bias, and who really controls the training pipeline.

One of the more surprising entries looks at how AI is turning the Iran conflict into theater. The idea that AI-generated content—deepfakes, synthetic narratives, automated propaganda—is being weaponized to shape public perception of a real-world conflict is both fascinating and deeply unsettling. It’s not just about bombs anymore. It’s about what people believe is happening.

I appreciate that MIT Tech Review isn’t sugarcoating this. The tone across these stories is sober, not breathless. They’re not cheerleading for defense contractors or panicking about killer robots. They’re documenting a reality where AI systems are becoming de facto advisors in military command structures, and the implications are still being worked out in real time.

If you’re a subscriber, you can grab the PDF or ePub versions. The related stories list includes a piece called “10 Things That Matter in AI Right Now: The new war room” and another on how the Iran conflict is being reshaped by AI-generated content. These aren’t sidebar reads—they’re central to understanding where defense strategy is headed.

What I find most striking is how quickly this went from experimental to operational. Five years ago, AI in the military meant better drone footage analysis or logistics optimization. Now we’re talking about chatbots advising on airstrikes. That’s a leap that should make everyone pause, regardless of their politics.

The eBook is behind a paywall, but the individual stories are worth tracking down if you don’t subscribe. This is one of those rare collections where every piece earns its place. No filler, no fluff—just hard reporting on a topic that’s too important to ignore.