Mining has always been an industry shaped by uncertainty. 

Finding orebodies in geologically complex settings, allocating capital with multi-decade implications, and operating assets across diverse regulatory environments are well-understood challenges.

What is changing is the context in which those decisions are made.

Discovery rates have declined over time, many orebodies are deeper and more complex, and expectations around safety, sustainability and capital efficiency continue to increase. At the same time, demand for minerals and metals is expected to grow as populations expand, infrastructure develops and the energy transition progresses.

As a sector, we need to continually improve how we discover, develop and operate assets – and it’s a big focus for BHP. This will increasingly be supported by how our people apply data, artificial intelligence (AI) and automation in their day-to-day work.

Execution matters more than invention

In mining, advantage rarely comes from the technology itself. It is shaped by how effectively it is applied to the problems and opportunities experienced by our teams, how well it is integrated into workflows, and whether it can be adopted at scale.

This places a strong emphasis on execution.

Mining is also highly interconnected. Decisions made in exploration can influence development. Constraints in processing can affect mine performance. Maintenance issues can have flow-on effects across the value chain.

At BHP, we see digital technologies as tools that can help our teams better understand these connections and make more informed decisions.

Building the data foundations that make AI work

We have extensive geological, operational, maintenance and safety data from more than 140 years of operations.

This includes everything from historical exploration records and field notes through to modern sensor, equipment and production datasets. However, much of this information has been collected over long periods of time, across different systems, formats and standards. This matters because AI is only as useful as the environment in which it operates. Where data is fragmented, even advanced tools will have limited impact. 

For this reason, a significant focus for us is on strengthening data foundations by improving standards, enabling interoperability and making data easier to connect and use in practice. This work is less visible than individual applications, but it is often what enables more consistent and scalable outcomes over time. 

Exploration provides a clear example.

Historically, a lot of geological information has been stored in formats that are difficult to search or integrate – from scanned reports to fragmented datasets accumulated over decades. In practice, this means geoscientists spend a considerable amount of their time finding, cleaning and reconciling data before interpretation can begin.

We have been developing a Central Data Platform to standardise and integrate exploration data globally. This is complemented by the use of digital tools - including generative AI - to help our geoscientists convert unstructured records into usable datasets.

In one example, work that previously took months to compile and interpret has been completed in hours using AI-enabled extraction tools. This means our geoscientists spend less time preparing data and more time analysing and testing ideas.

In a discipline largely characterised by uncertainty and long investment horizons, that distinction is important. These tools can help our teams test ideas faster, build confidence earlier and make better-informed decisions about where to allocate capital.  

Applying digital tools across BHP’s operations

The same principles apply in our operations, where we’re always seeking to maximise what we can produce from our mines.

At Escondida, our teams are using digital twins and AI models to better understand how ore characteristics influence processing performance. This supports more informed blasting and blending decisions before material reaches the plant. The approach has reduced losses associated with ore variability at the Semi-Autogenous Grinding (SAG) mill, with some production losses linked to granulometry reduced by around 70 per cent in targeted applications. 

Across the value chain, digital tools are also used to support predictive maintenance, monitor equipment in real time and identify potential issues earlier. 

At BHP’s Western Australia Iron Ore (WAIO) operations, computer vision is used at key points along conveyors to help our teams spot oversized rocks or foreign objects. This helps teams remove them before they create safety risks, damage equipment or cause unplanned stoppages. In pilot applications, this solution has been associated with reductions in disruption events, which historically contributed to over 1,000 hours of downtime across the system. 

These are practical applications that support our teams in managing real operational constraints.

Collaboration and scale

Mining is a large and complex system, and decisions in one part of the value chain often influence outcomes elsewhere. 

Better connected digital environments can help miners better understand the knock-on effects and plan in a more coordinated way. This starts with strong data foundations, safe and disciplined implementation, and connecting solutions into broader systems rather than isolated tools.

Data fragmentation, interoperability and standards are not unique to BHP. This creates opportunities for collaboration in areas that are not commercially sensitive, and where alignment may improve efficiency and shared learning.

BHP’s Xplor program is one example, bringing together early-stage exploration companies, geoscientists and technology teams to support new approaches to discovery.

In addition, BHP is working extensively with governments, industry and research institutions to support collaboration and innovation across the sector. 

In February 2026, BHP entered into a collaboration with South Africa’s Council for Geoscience to help unlock legacy datasets. 

And in March 2026, BHP took a major step to partner with the Government of Canada in support of the digitisation of Canada’s extensive geoscience and drill core data. 

Over the past four financial years, our digital and analytics initiatives have delivered more than US$2 billion in value for BHP. That value has spanned many parts of the value chain – from exploration to operations to logistics – but we have barely scratched the surface and we’re looking for partners who can help us dig deeper.  

No company can solve data fragmentation or system integration alone. BHP is very open to working with others in this space because we believe being smart about what is shared allows everyone to make better decisions.

Outlook

It is clear the world will require more resources over the coming decades – and meeting that demand responsibly will require our industry to get even better at discovering, developing and operating mines.

AI is going to play an increasingly important role in delivering the resources the world needs right across the value chain.

While individual applications can deliver targeted improvements, the larger AI opportunity 
lies in connecting them more effectively - linking data, models and decisions across exploration, development, operations, maintenance and processing.

In an industry as complex as mining, improving decision-making across the value chain will remain an important source of value. 

Importantly, these technologies do not replace the expertise of our people. They support people to apply their judgement and experience more effectively.

Mining has always been about managing uncertainty – finding ways to achieve the seemingly impossible. 

The good news is that AI is improving our ability to navigate that uncertainty with greater efficiency, insight and discipline – to make possible tomorrow what seems impossible today.