Datacenters Are About to Bury Us in Metal-Rich Waste

By Siam Sukkhee Trading Co., Ltd — 2026-09-18 — Waste360 (recycling/scrap)

There's a new mountain coming. Nobody's really talking about it yet, but it's there in the data.

The Basel Action Network dropped a report last week—actually the first part of a four-part white paper—that puts a number on something most people would rather ignore. When you build thousands of massive datacenters to run AI systems, the hardware doesn't last forever. It gets hot, it fails, it becomes obsolete in five years or less. And then it becomes waste.

Here's where it gets interesting for us.

The report estimates that AI datacenters alone will generate between 395 and 617 million tonnes of electronic equipment waste between now and 2050. That's forty to sixty times higher than earlier estimates. To visualize it: enough to fill 15 to 23 million shipping containers. Six times around the globe, according to another way the report frames it.

Right now, the world is on track to triple its annual generation of toxic e-waste over the next 25 years, more or less. That's the background noise. The datacenters are the amplifier.

E-waste contains the metals we work with every day. Copper, obviously. Zinc, cadmium, nickel, tin. Also gold, silver, platinum in smaller quantities. And toxins—lead, mercury, cadmium again. That's the thing about electronics. The valuable materials and the poisonous ones are sitting right next to each other in the circuit boards.

The problem is extraction.

Vince Beiser, who wrote a book on the subject called Power Metal, said something that stuck with me: "We're doing a terrible job of handling e-waste in the US. Some of it gets dumped overseas. Most of it just gets dumped in landfills, which is a problem because it can leach all kinds of nasty things into the water, into the soil." He's not wrong. Only about a fifth of electronic waste globally gets properly recycled and recovered.

Actually, that's not quite right either. The numbers I've seen range from one-fifth to 22 percent depending on the source. Neither is good. The point is the same: most of it doesn't get processed. It sits. It leaches. Communities near disposal sites get contaminated.

This is where the conversation should shift. In Thailand, in Southeast Asia, we're not far removed from the effects of this. We know what happens when waste ends up in the wrong places. And we know what metals are worth recovering.

The Basel Action Network report doesn't just warn about volume. It also flags that there's no coordinated plan. No infrastructure. No standards. The recycling industry exists, sure, but it's fragmented and under-scaled. Equipment lifespans are shortening—five to seven years for liquid-cooled systems. The machinery that processes e-waste is energy-intensive. Separating materials safely requires specialist knowledge and equipment that most places don't have.

For a trading company like ours, this creates an odd situation. There's going to be more material in circulation, more pressure to recover metals from waste streams. But the systems to do that extraction profitably and responsibly aren't there yet. Not at scale.

The opportunity isn't in the next two years. It's in what comes after—when someone finally decides this waste needs to be handled properly, and those who figured out how to do it efficiently will matter.

Source: "Already electronic waste is the fastest-growing waste stream on the planet and only one-fifth is properly managed" — beSpacific

Tags: copper market Southeast Asia