AI industry demand drives ruthenium prices to a record high and raises the risk of a shortage

Ruthenium, a metal used in hard drive manufacturing, has climbed to an all-time price high. The main driver is fast-growing demand from AI infrastructure, while supply remains tight and may fail to keep up with the market as early as 2026.

Why demand went up

This is an unusual situation for the commodities market: a metal that sits outside the mainstream conversation suddenly becomes critical to an entire technology supply chain. That is exactly what is happening with ruthenium now. As companies build new data centers and expand storage for training and serving AI models, demand has grown for the components this infrastructure cannot scale without. Attention usually goes to chips and accelerators, but physical data storage remains an essential part of the AI economy.

That is bad news for the industry, because the demand growth looks less like a one-off spike and more like the continuation of a longer trend. Companies investing in AI services are simultaneously expanding compute, redundancy, and archives, which means they are buying more drives and related materials. When one of those materials runs short, the pressure moves quickly down the chain: from component makers to system integrators and then to large enterprise customers.

Where the shortage risk comes from

The problem is that ruthenium is nearly impossible to bring to market quickly in large volumes. It is not mined as a standalone metal but recovered as a byproduct of mining other platinum group metals. That makes supply inflexible: even if the price jumps sharply, mining companies cannot simply open dedicated capacity for ruthenium. Geography adds another risk. The market depends heavily on South Africa, and any delay, disruption, or drop in output there immediately affects the global balance.

Because of how the market is structured, a shortage may arrive not gradually but all at once. If demand from AI and data storage holds up and supply stays at current levels, even a small amount of additional consumption could push the price sharply higher. Unlike larger commodity markets, this one has almost no slack: if industry needs the metal right now, there is simply nothing to quickly close the gap between supply and demand.

• ruthenium production cannot be ramped up quickly as a standalone mining operation
• key volumes are tied to a single region
• drive makers will compete for limited raw material
• price increases quickly feed into equipment costs

Who will pay more

Hard drive makers and storage system vendors will feel the pressure first, especially those serving large data centers. For them, the rising price of ruthenium is not an abstract commodities story but a risk of higher costs for specific components and squeezed margins. If vendors cannot pass the higher costs on to customers right away, some contracts will become less profitable. If they can, the costs will move further up the chain into the price of server infrastructure, backup storage, and large-scale AI projects.

That does not mean the entire storage market will instantly get more expensive across the board. But if the shortage does materialize as early as 2026, companies will start acting ahead of it: revisiting purchasing plans, locking in volumes through long-term contracts, and running the numbers on storage economics more carefully. For businesses building AI products, it is one more reminder that infrastructure costs are not set by GPU prices alone. A bottleneck can emerge in a less visible material and hit the budget unexpectedly.

What it means

The ruthenium story shows how much the AI boom depends on raw materials that rarely come up outside the industry. When demand for compute grows faster than the mining of rare materials, the risk shifts from labs and models to the entire supply chain, from storage drives to data center economics.