Can a Radioactive Ore Break China’s Grip on Rare Earths?

If you’re building anything in hardware right now—from the servers running LLMs to the robots designed to replace manual labor—you are, by proxy, a customer of the Chinese state. For years, the narrative in the West has been one of "de-risking," a polite way of saying we’re terrified that 90% of the rare-earth oxides and metals needed for modern life come from a single, increasingly grumpy supplier.

We aren’t just talking about a few magnets in your iPhone. Rare earths are the oxygen of the high-tech economy. They power the permanent magnets in every electric vehicle motor, every tactical drone, and every high-performance cooling system in a data center. Without them, the "AI revolution" stays trapped in a PowerPoint presentation. But the hunt for a domestic alternative has led us to a mineral called monazite, and it’s a perfect case study in why founder-led hardware is so incredibly difficult.

The Monazite Gamble

For decades, monazite was the industry’s black sheep. It’s an ore that is packed with the stuff builders actually want: neodymium, praseodymium, and dysprosium. These are the "magnet metals" that make electric motors efficient. On paper, monazite is superior to the more common bastnaesite ore because it’s easier to find in "heavy mineral sands"—literally dredging up beaches rather than blasting through mountains.

But there’s a catch that has kept most sane founders away: thorium. Monazite is radioactive. When you process it to get the rare earths, you’re left with piles of thorium and uranium. In the eyes of a regulator, you aren’t just a mining company anymore; you’re a nuclear waste manager. And that is where the business model usually falls apart.

From a founder’s perspective, this is the ultimate regulatory moat. It’s why Energy Fuels, a company based in Utah that already operates a uranium mill, has suddenly become the most important player in the U.S. rare-earth space. They already have the licenses to handle the "hot" stuff. They aren't trying to build a new regulatory framework from scratch; they’re just adding a new product line to their existing waste-management infrastructure. That’s the kind of pragmatic, builder-first pivot I respect.

The "Afghanistan of Mining"

Despite the hype, the path to rare-earth independence is littered with the carcasses of companies that thought they could "disrupt" geology. I’ve heard the industry called the "Afghanistan of Mining"—a place where grand ideas go to die in a quagmire of shifting prices and technical failures.

Look at the Saskatchewan Research Council (SRC) in Canada. They’ve spent over $100 million on a processing plant that was supposed to be pumping out 40 tonnes of rare earths a month by now. Instead, they’ve produced almost nothing. They’re stuck in a classic founder trap: they built the refinery before securing the feedstock. They thought the monazite would just show up, but export bans in places like Vietnam and the sheer cost of building infrastructure to local mines have left them running "sample runs" while the clock ticks on their funding.

This is a warning to anyone building in the physical world. Your tech can be perfect, your chemistry can be revolutionary, but if your supply chain is theoretical, your business is a ghost. In the rare-earth world, speculators love to "pump and dump" stocks based on a promising drill hole, but very few actually have the stomach to build the separation facilities required to make a sellable product.

The Waste-to-Fuel Opportunity

Here’s where it gets interesting for the AI and energy crowd. If we actually commit to monazite, we’re going to produce thousands of tons of thorium. Right now, that’s a liability. It’s waste that has to be buried in lined ponds for eternity.

But for a skeptical optimist, thorium is the holy grail of carbon-free energy. One tonne of thorium can generate about 350 megawatts of electricity. If we scaled up monazite processing to meet our magnet needs, we’d produce enough thorium to power half of the projected growth of the entire U.S. grid. We have startups like Copenhagen Atomics and Thorizon trying to build molten-salt reactors to use this "waste" as fuel.

If I were a founder looking at the next ten years, I wouldn't just look at rare earths as a mining play. I’d look at the convergence of three things: the desperate need for magnets, the regulatory necessity of handling radioactive waste, and the massive energy demands of AI data centers. The person who connects those three dots wins the decade.

A Reality Check for Builders

We need to stop pretending that "breaking China’s grip" is just about writing a check to a mining company. It requires a fundamental shift in how we handle industrial regulations. As long as thorium is treated solely as a toxic liability rather than a potential energy asset, the costs of domestic rare-earth production will remain prohibitively high.

For those of us in the trenches of AI and crypto, this might seem like a distant problem. It isn't. Every GPU you buy and every robot you deploy depends on this fragile, radioactive supply chain. The current U.S. strategy of throwing billions at speculative projects—some of which seem to have more to do with political connections than engineering—isn't working.

We need a builder-first approach that prioritizes existing infrastructure, like the White Mesa Mill in Utah, and stops chasing "miracle" deposits in scenic areas where a mine will never be permitted. The future of monazite is real, but only if we’re honest about the radioactive baggage that comes with it.

The takeaway: Don’t bet on the company with the biggest hole in the ground. Bet on the company that knows how to navigate the waste. In the rare-earth war, the winner won't be the best miner; it'll be the best chemist who can survive a government audit.

Read the original at IEEE Spectrum →