AI Boom Puts New Focus On Mining And America’s Critical Mineral Supply
AUSTIN, Texas (TXAN 24) — Artificial intelligence is often discussed in terms of software, chips, data centers and computing power. But one mining executive at the Reuters Momentum AI event in Austin says there is another part of the AI economy that deserves considerably more attention: the mines supplying the metals needed to build it.
Paul, an executive with Americas Gold and Silver, told TXAN 24 News that the rapid expansion of artificial intelligence is increasing pressure on an already constrained global supply of metals such as silver, copper and antimony.
“AI is not in the clouds. AI comes from the mines,” Paul said.
Americas Gold and Silver operates three mines in Idaho and another operation in Mexico. Paul said the company employs approximately 400 miners in Idaho, where its operations are underground rather than open pit.
The company mines rock, transports it to the surface through a shaft and processes it through a mill where the material is crushed and ground. The resulting concentrate is used to produce five metals: silver, antimony, lead, copper and gold.
Paul said the company’s Idaho operations are fully permitted and already producing domestically, which he described as an important advantage as the United States looks to strengthen domestic supply chains.
The connection between artificial intelligence and mining can be easy to overlook.
The growth of AI has fueled construction of large data centers, increased demand for semiconductors and expanded investment in electrical infrastructure. Those systems require significant quantities of physical materials, including copper and silver.
Paul said the issue is particularly important because mineral production cannot expand at the same pace as technology development.
“You look at the speed in which AI is forming,” he said. “It takes us anywhere from 15 to 20 years before we have a discovery till we build the mine, till we permit it.”
That timeline creates a potential supply-chain challenge.
While AI companies can rapidly expand computing capacity, developing a new mine involves exploration, discovery, engineering, permitting, construction and ultimately production.
Paul said that difference in timelines is one reason he believes the mining industry needs to be part of conversations taking place around AI infrastructure.
Silver has traditionally been associated with jewelry, coins and bullion, but Paul said its industrial applications have become increasingly important.
Silver is used in a range of electronics and industrial applications, including components associated with modern technology.
Paul said silver therefore occupies two roles in the global economy: a precious metal and an industrial metal.
“When you start coming to these conferences, what you learn really, really quickly is some of the challenges that the AI industry is facing,” Paul said.
He argued that increasing demand from AI, electrification, solar energy, electric vehicles and consumer electronics could place additional pressure on silver supplies.
Paul said his company expects to produce approximately 3.2 million to 3.6 million ounces of silver this year, along with copper, antimony, lead and smaller amounts of gold.
He also said global silver production has fallen short of demand, including recycling, for six consecutive years.
One of the central challenges Paul described is that higher commodity prices do not automatically create new production.
Mining companies still have to find deposits, develop them, secure permits, build infrastructure and begin operations.
“You can’t build and develop a mine that fast,” Paul said.
He compared the development timeline of a mine with the rapid growth of artificial intelligence, saying technology is advancing at a pace that mining companies cannot immediately match.
Even an existing mine cannot simply double or triple production overnight.
“You don’t go in, flip a light switch and go, all right, we’re tripling production in a year,” Paul said.
Instead, increasing production can require additional infrastructure, new mining methods, additional equipment, permitting and capital investment.
Paul said his company is attempting to address that challenge by investing heavily in its existing operations.
The company is revitalizing a mining district in Idaho with a history stretching back roughly a century. Paul said the broader Silver Valley has produced more than 1.3 billion ounces of silver historically.
He said the company has raised nearly $400 million over the past year and a half, with approximately 90% to 95% of that capital deployed toward its operations in Idaho and Mexico.
The investments are being used for infrastructure, modernization and production growth.
Paul said the company expects this year to be a record year and anticipates surpassing that production record next year.
While silver and copper are important to technology and industrial applications, Paul also emphasized antimony.
Antimony is used in several applications, including flame retardants and certain military-related technologies.
Paul said approximately 85% of the world’s antimony production comes from China and noted that China has previously restricted or stopped exports of the material to the United States.
That dependence, he said, creates a vulnerability for American supply chains.
“85% comes from one country, China,” Paul said. “So what happens then? Do we just stop producing? We can’t do that.”
His argument is that domestic production of critical minerals could reduce America’s reliance on foreign sources while providing materials for industries considered strategically important.
Americas Gold and Silver is also looking toward additional domestic processing capacity.
Paul said the company is planning a new antimony facility in Idaho, which he described as the first such facility in the United States in a very long time.
Construction is expected to begin in 2027, according to Paul.
The proposed facility would allow the company to further process antimony produced from its domestic mining operations.
That would place the company further along the domestic supply chain, from exploration and mining through processing.
Paul described his company as being positioned near the beginning of the supply chain.
A mineral deposit is identified through drilling. The company then models the discovery, extracts the rock, transports it to the surface, processes the material and produces the resulting metals.
“We’re one of the only domestic from beginning to end,” Paul said.
He said the company’s goal is to continue expanding production while modernizing its existing operations.
The broader issue, however, extends beyond one company.
As the United States builds more data centers and expands its electrical infrastructure, the demand for the physical materials behind those projects could become increasingly important.
Paul said he attended the Austin AI event in part to listen to technology executives and better understand the industry’s growth. He said he was surprised that his company was the only mining company represented at the event, given the industry’s dependence on raw materials.
Paul pointed to the rapid expansion of data centers as one example of why the issue matters.
He cited thousands of data centers already operating in the United States, along with plans for additional facilities.
Those projects require electrical systems, semiconductor components, power infrastructure and other equipment — all of which ultimately depend on physical materials.
For Paul, the discussion about AI therefore cannot stop at software and computing.
The supply chain begins much earlier.
“It starts at the drill hole,” he said.
Without the metals, he argued, the infrastructure required to expand artificial intelligence cannot be built at the same scale.
Another issue Paul raised was the length of time required to permit and develop mines in the United States.
He said domestic production needs to be supported by permitting processes that allow projects to move forward more efficiently while maintaining appropriate environmental and regulatory standards.
Paul also emphasized the relatively small surface footprint of underground mining compared with open-pit operations.
His broader argument is that the United States needs to balance environmental concerns with the growing demand for critical minerals.
“We need to work together. We need to work in alignment,” he said.
Looking ahead, Paul said the central question is whether global mineral production can keep pace with continued technological growth.
If AI development, electrification, solar deployment and other industrial trends continue expanding, he said the gap between mineral demand and supply could widen.
That could put additional pressure on commodity prices and increase the importance of companies that are already producing.
Paul said higher prices alone cannot solve the problem because developing new mines takes years.
He characterized the current environment as the early stages of what he believes could become a major period for silver.
He also discussed the possibility of substantially higher silver and gold prices, though those figures represent his own outlook rather than guaranteed future prices.
The conversation in Austin highlighted an often-overlooked component of the artificial intelligence economy.
AI may operate digitally, but the infrastructure supporting it is physical.
Data centers require power. Power systems require infrastructure. Semiconductors require materials. Electronics require metals. And those metals ultimately have to come from somewhere.
For Paul, that makes mining part of the AI conversation—not separate from it.
As the United States continues investing in artificial intelligence, the question may increasingly become not only how quickly the technology can advance, but whether the underlying supply chains can keep pace.
“Without a product, we ain’t building AI centers,” Paul said.
For the mining industry, the race surrounding artificial intelligence could therefore become another race entirely: finding, producing and processing enough of the raw materials needed to build the technology’s physical foundation.
Comments