BHP is using artificial intelligence and advanced computing to accelerate the search for new ways to extract copper from ore, as the miner looks to improve recovery rates across its global copper portfolio.
The company is working with Microsoft and Prescience Insilico on a project that has screened more than 500,000 molecules that could improve copper leaching, a process used to recover copper from ore.
Using Microsoft’s Discovery platform, the project completed tens of thousands of simulations and quantum chemistry calculations, narrowing the field to a shortlist of candidates now being tested in Australian laboratories.
“As copper demand grows and new deposits become harder and more expensive to develop, improving recovery from existing ores is a critical lever to help meet future supply needs,” BHP vice president innovation Jessica Farrell said.
Copper is central to electrification, digital technologies and AI, placing greater pressure on miners to improve efficiency as future demand rises. But finding new ways to improve copper recovery has traditionally relied on laboratory testing and trial-and-error processes, with millions of potential molecules making the task slow and complex.
“This partnership has given our technical experts the tools they need to narrow an almost infinite field of possibilities down to a small number of options that could one day be deployed in our global copper operations,” Farrell said.
“Those candidates are being tested against the realities of our orebodies and operating constraints, so we are solving for what can actually work in practice.”
For BHP, the project sits within a broader strategy that places copper, Tier-1 assets and operational discipline at the centre of the company’s next phase of growth.
BHP incoming chief executive officer (CEO) Brandon Craig recently used his first major address since taking over from Mike Henry to frame the miner’s future around continuity rather than reinvention.
“BHP has a clear and compelling strategy: to invest in highly attractive and durable commodities; operate world-class Tier-1 assets with excellence; offer a distinctive approach to social value; and allocate capital with discipline,” Craig said at the BofA Metals, Mining and Steel Conference 2026.
BHP is currently the world’s largest copper producer, with major exposure through Escondida in northern Chile and Copper South Australia, while the emerging Resolution Copper project in the US has reached a major land exchange milestone, enabling further development work.
The company is targeting copper production of around two million tonnes per annum by 2035, supported by a pipeline of organic projects expected to deliver 3–4 per cent compound annual copper-equivalent growth. Beyond 2035, Craig has also highlighted “programmatic growth” through exploration, partnerships and selective bolt-on acquisitions.
BHP’s AI-led leaching work points to a key part of the copper growth equation, one that involves extracting more value from known resources, rather than relying solely on discoveries or major acquisitions.
That challenge’s cruciality is growing as new copper supply remains difficult to bring online due to permitting delays, capital intensity and declining grades across mature mining jurisdictions. As demand rises, the ability to improve recovery from existing ores could become a major lever for large producers.
The technical challenge is particularly clear in relation to chalcopyrite, the mineral responsible for roughly 70 per cent of global copper supply.
Research from Monash University, published in Nature Geoscience, has revisited chalcopyrite and highlighted how its structural complexity could influence future processing routes.
Despite being the world’s dominant copper-bearing mineral, chalcopyrite has historically been difficult to process through low-temperature leaching. This has limited the viability of some hydrometallurgical approaches and kept many operations reliant on energy-intensive pyrometallurgical routes such as smelting.
“Chalcopyrite is the world’s primary copper mineral, but it behaves in surprisingly complex ways that have limited how efficiently we can extract copper from it,” study lead Professor Joël Brugger said.
The Monash research found that chalcopyrite’s crystal structure is not uniform, with microscopic defects and trace elements, including silver, gold and nickel, influencing how the mineral behaves during processing.
One of the study’s key findings was the role of trace silver, which can destabilise the mineral’s surface and trigger reaction cycles that enhance copper release.
“By understanding how trace elements like silver interact with chalcopyrite at the atomic level, we can begin to design smarter, more targeted extraction methods,” co-author Barbara Etschmann said.
“That means less energy, fewer chemicals, and better recovery from the same resource.”
BHP’s AI project, its broader copper strategy and the Monash research point to future copper growth relying on finding and developing new deposits, but also on processing existing resources more intelligently.
This could make recovery improvement one of the most important frontiers in copper; if AI can shorten the search for viable processing solutions and mineral science can better explain why certain ores resist extraction, it grants miners the tools to lift output while reducing energy use, chemical intensity, and operating costs.
