Critical Minerals and the New Great Game: What's Actually at Stake

China processes 87% of the world's rare earth elements. Not mines — processes. That single number explains more about the next decade of geopolitics than most foreign policy essays I've read this year.

And yet most people I talk to — even sharp ones running supply chains for a living — still think "critical minerals" means lithium for Teslas. It's a much bigger story than that.

Let me try to lay out what's actually going on, because the noise around this topic has gotten loud and the signal keeps getting buried.

The list keeps growing, and that's the point

The US Geological Survey now tracks 50 critical minerals. The EU lists 34. Japan has its own. India quietly published one in 2023 with 30 entries. These lists don't fully overlap, which tells you something — every country defines "critical" based on what it can't get and what it desperately needs.

Gallium. Germanium. Graphite. Niobium. Tungsten. Cobalt. Nickel. Manganese. Dysprosium and neodymium for the magnets inside every EV motor and wind turbine. Antimony for ammunition primers (yes, really). Indium for the touchscreen you're probably reading this on.

Here's the thing most coverage misses: the bottleneck usually isn't extraction. It's processing. The DRC produces around 74% of global cobalt, but most of it ships to China for refining. Indonesia mines nickel; China builds the smelters. Australia digs up lithium spodumene; it travels to Jiangxi to become battery-grade lithium hydroxide.

So when politicians announce a new mine in Nevada or Greenland, I half-shrug. A mine without a refinery is just a hole in the ground that exports raw material to your competitor.

Why exploration suddenly matters again

For about 30 years, mineral exploration was a backwater. Junior miners in Toronto and Perth kept the lights on, but real money — venture, sovereign, strategic — went elsewhere. That's flipped hard since 2021.

The US Defense Production Act has been invoked for cobalt, nickel, graphite, manganese, and lithium. The EU's Critical Raw Materials Act sets a 10% domestic extraction target by 2030. Saudi Arabia is throwing billions at Manara Minerals. India's KABIL is doing deals in Argentina. Japan's JOGMEC has been at this quietly for years and looks prescient now.

But the actual exploration work? Still painfully slow. The average mine takes 16 years from discovery to first production in the West. Sixteen years. You can't catch up to a 30-year head start in 16-year increments.

This is why satellite-based spectral analysis is having a moment. Companies like GeoMine AI are using hyperspectral satellite data to identify mineral signatures across huge swaths of terrain in weeks instead of years — work that previously needed boots, helicopters, and a lot of luck. It doesn't replace drilling. Nothing replaces drilling. But it dramatically narrows where you drill, and in a world where capital wants answers in quarters, not decades, that compression matters.

I used to think the bottleneck in critical minerals was political will. After spending more time with geologists this past year, I think it's actually data. We don't know what's underneath most of Africa, central Asia, parts of South America, or — surprisingly — large stretches of Pakistan and Iran. The maps are old. The surveys were done with 1970s instruments. We're flying blind on a continent of opportunity.

The decoupling that isn't really decoupling

Here's where I'll probably annoy some readers. The Western narrative around "de-risking" from Chinese mineral supply is mostly theater right now. The numbers don't lie.

In 2023, US lithium-ion battery imports from China grew 41%. EU rare earth dependency on China stayed above 95%. Even "friendshoring" deals — Indonesia, Morocco, Chile — often involve Chinese capital, Chinese engineering, or Chinese offtake agreements buried in the fine print.

Look, I'm not saying the strategy is wrong. Diversification is the right move. But the timelines being floated by Western governments (5 years, 7 years) are fantasy. The honest number is closer to 15 to 20 years to build genuinely independent supply chains for, say, neodymium magnets or battery-grade graphite.

Meanwhile China isn't standing still. They've started restricting exports of gallium, germanium, graphite, and certain rare earth processing technologies. Not banning — restricting, which is more sophisticated. It signals capability without triggering the WTO. And it forces every Western planner to assume worst-case scenarios when modeling capex for the next plant.

The countries that win this won't be the ones with the loudest industrial policy. They'll be the ones who figured out three boring things: where the deposits actually are, who's allowed to process them domestically, and how to finance a 12-year project when interest rates won't sit still.

What this means if you're not a mining executive

If you run a manufacturing business, your bill of materials has more geopolitical exposure than your CFO probably realizes. Permanent magnets, semiconductors, batteries, specialty steels — all of these have at least one input where a single country controls more than 60% of refined supply.

If you're an investor, the public junior mining sector is mispriced in both directions — some names trade like they've already found a tier-1 deposit, others trade like they'll never raise another dollar. Picking is brutal. The royalty companies and the picks-and-shovels plays (drilling, geophysics, software) tend to be the saner exposure.

If you're a policymaker reading this in Islamabad, Jakarta, Lusaka, or Lima — your country is sitting on assets the world will pay almost any price for in 10 years. The question is whether you'll capture the value or watch it ship out as concentrate, the way it has for a century.

So what is actually at stake? Not just EVs. Not just defense. The whole industrial stack of the 21st century — energy, computing, mobility, weapons — runs on roughly two dozen elements, and we've spent a generation pretending the supply of those elements was someone else's problem.

It isn't anymore. The question now is who's paying attention early enough to do something useful about it.