With accelerating momentum in Artificial Intelligence (AI), its impact can be seen not just in new data centers, but in the rising global demand for metals. Examining what U.S. metals flows can reveal about advancements in AI was the focus of a panel discussion that Helen Amos, Macro Commodity Analyst, BMO Capital Markets, moderated at the 35th Annual BMO Global Metals, Mining & Critical Minerals Conference.

To explore how AI is shaping the demand outlook for metals, the following speakers joined the discussion on stage:

• 

  Kristin Milchanowski, Chief AI & Data Officer, BMO Financial Group

• 

  Lief Williams, Managing Director, AI Investment, North America, Brookfield Infrastructure Partners

• 

  Jin Hennig, Managing Director & Global Head of Metals, CME Group

• 

  Christy Wright, EVP of Supply Chain, Interim CFO, Southwire Company

Per the panel, there is a growing appreciation for the importance of mining and metals to AI infrastructure and what they mean for the global economy. As noted by Kristin Milchanowski, AI is accelerating the capital expenditure supercycle, where companies are pouring money into physical assets such as data centers and power grids.

Amid the shake-up, the metals industry is coming into sharper focus. Copper, in particular, is attracting attention because of the build-out of data centers, power generation, transmission and grid upgrades required to keep up with AI development. “What we’re experiencing is not a short-cycle tech trade,” said Milchanowski. “AI demand is generating a real, long-duration infrastructure-driven demand for copper.”

Jin Hennig of CME Group, which operates one of the world’s largest commodities exchanges, said growing demand for metals has attracted new investors. She noted three trends that have emerged over the past 18 months:

• 

  Institutional investors are increasingly reassessing their views on certain metals. One example she highlighted was the growing divergence between oil and copper prices, which historically have moved in tandem. More recently, copper prices have risen even as oil prices have softened, reflecting a shift in how institutional investors view copper’s role, Hennig said.

• 

  The second is that metals have become more mainstream. “People tend to get more familiar with the things that they hear on the news,” she said, pointing to silver as an example where talk about rising demand for the precious metal spurred retail investor participation.

• 

  Lastly, Hennig said she is seeing cryptocurrency investors entering the precious metals space as they seek other assets to anchor their portfolios.

While demand for metals and investor interest have increased, some governments are trying to stockpile critical minerals, including copper.

While sourcing raw materials to build the data centers behind the AI revolution is one challenge, powering that infrastructure is another, said Brookfield Infrastructure Partners’ Lief Williams. A growing number of new data center developments now involve multi‑hundred‑megawatt commitments, and many of these sites are being selected specifically for their ability to scale over time; from initial loads of around one gigawatt to several gigawatts, and in some cases potentially five to 10 gigawatts over the next decade, he said.

The most immediate challenge in addressing this issue will likely come from available capital and labor, he said, adding that the physical constraints would take more time to solve.

That’s consistent with what Christy Wright of Southwire, one of North America’s largest power cable manufacturers, said about demand for materials needed to address those power constraints. “There are strong growth prospects in power generation,” she said. “We serve a broad range of markets and actively manage our supply chain. As a result, we are not seeing shortages in the materials needed to support our customers."

All the panelists agreed that governments will be integral in developing the infrastructure required to keep up with AI demand. Milchanowski noted that many critical components of power infrastructure, such as access to land, water, and permitting, are controlled by governments, meaning public authorities could play a central role in enabling data center development alongside hyperscalers.

Williams added that governments, including the U.S. and China, are very much supportive of AI infrastructure development on both the demand and supply side. Other governments are also taking notice of the disruptive potential of AI on national economies.

Wright agreed, stating that government involvement will be essential in supporting the permitting processes to alleviate bottlenecks.

Like the rest of the panelists, Hennig said she, too, is hopeful that governments will step up to help take some of the burden off the private sector as the global AI race heats up. “I feel very optimistic,” she said. “There are big problems, but I think the willingness, the funding is there – and I think that’s great news for the industry.”