Data: The hottest non-commodity in the mining & metals sector?

There is a tendency to think of data as generic, especially when it is vast, but the uniqueness of mineral data makes it the ultimate non-commodity, primed to take advantage of AI.

We previously explored whether technology was becoming the hottest commodity in the mining & metals sector. While technological applications in the sector take many forms, generative artificial intelligence (GenAI) may have the most exciting upside across the sector, from exploration to operations. However, the application of AI, like many products in the sector, is only as good as its feedstock: data. Here, we examine whether data has become the hottest non-commodity in the mining & metals sector.

Teaching the technology

The hype around AI is real. In particular, AI's application to mineral exploration could potentially power the technological revolution needed for the mining & metals sector to achieve the productivity growth that would allow it to meet the demands of the energy transition. However, for the sector to use AI effectively, AI must learn. It does that by being fed existing data.

To understand how AI uses data, consider how children are taught. GenAI and machine learning algorithms need examples to learn from. Similarly, children learn to recognize animals by looking at pictures and being told their names. In the case of AI, the data it is fed serves as the pictures and names. The more data that AI is given, the better it becomes at identifying patterns, making predictions and generating insights. In the context of mineral exploration, AI's ability to detect geological anomalies that warrant further studies requires that AI understands what these anomalies look like so it can effectively sift through data sources to identify them more rapidly.

Using substantial stores of data

The mining & metals sector is already using data to drive the application of AI and machine learning.

For example, KoBold Metals, a new entrant into the sector, uses AI to guide exploration for deposits of critical battery metals. KoBold's AI employs a proprietary data set—comprising geological data compiled from multiple sources—to identify promising areas for further exploration, significantly reducing the time and cost associated with the exploration process. Similarly, Computational Geosciences Inc., a 94 per cent-owned subsidiary of Ivanhoe Electric, Inc., used its proprietary software to rapidly convert high volumes of raw data gathered from Ivanhoe Electric, Inc.'s proprietary Typhoon™ technology to generate 3D models of underground geology that identify geophysical anomalies for follow-up drilling. These examples highlight that, while AI and machine learning algorithms can process and interpret data at unprecedented speeds, their effectiveness hinges critically on the quality and comprehensiveness of the data they are fed.

However, the new kids on the block aren't the only ones applying technology to accelerate exploration efforts. Sector stalwarts with vast core libraries and historical data sets spanning generations are also reaping the benefits of applying AI and machine learning. For example, Rio Tinto has implemented AI and machine learning to automate core sample logging and improve orebody modelling, leading to more efficient exploration and resource estimation. These technologies sift through years of accumulated data to provide more accurate resource assessments and reduce the time to discovery. Similarly, BHP leverages AI to integrate core sample data with other geological information, improving its understanding of ore bodies and optimizing its mining operations. And let's also remember the core libraries and historical data of the major oil & gas companies, which in many cases completely overshadow those of the mining companies—all of these can be re-examined with the new aim of locating energy transition minerals.

The growing value of data

Given these applications, we have seen the value of data grow in the sector. KoBold Metal's November 2023 acquisition, from Alaska Energy Metals Corporation, of exploration data neighbouring its Skolai Project demonstrates the value that is being ascribed to data sets in an asset to be used as part of a broader application of AI.

In addition to straight acquisitions, we are seeing more new strategic partnerships between traditional and non-traditional participants in the mining & metals sector. These innovative partnerships benefit from the combination of traditional mining majors' expertise and data sets, collated over years of mining and exploration activities, and the new sector entrants' technology, which can analyse those data sets. By joining forces, traditional and non-traditional sector participants are unlocking their unharnessed potential.

Increased transactional activity related to data raises novel questions that historically may not have garnered as much attention in a typical transaction in the mining & metals sector. For instance:

1. How is the data's value determined? Often the answer is on a case-by-case basis given the unique nature of the data and the context in which it is being acquired or, in the case of a strategic partnership, made available to the joint venture.
2. Who owns the data? In the context of an acquisition, does the "seller" retain any rights to use the data? In a strategic partnership, does ownership reside with the contributor of that data? What happens if that partnership breaks up, and who owns the resulting intellectual property generated by the data?

Increasing governmental control?

However, underlying these high-profile private transactions, natural resources are generally considered a nation's wealth, which begs the question, whose data is it anyway? As technological innovation continues to permeate the sector, we may see more governmental control not only as to the end-products of mining & metals activity but also as to the data that informs that activity in the first place.

In certain jurisdictions, exploration results are owned by the country in which they are found, rather than by the company or person that gathered them. Certain countries which have large reserves of natural resources, such as Australia, Canada and Brazil, require mining companies to report the exploration data they collect to government agencies. This data is often made publicly available after a certain period.

Consequently, in jurisdictions where data is reported to governmental agencies and otherwise made publicly available, partnerships with governments could become key enablers of this next frontier of mineral exploration.

For example, in Zambia, where the government has recently announced a country-wide airborne geophysical survey, Ivanhoe Mines is partnering with the government to share existing and new geological data, for the purpose of co-developing prospective projects.

Unlocking GenAI's potential for mineral data

As the mining & metals sector increasingly leverages AI and other advanced technologies, the importance of data cannot be overstated. Mining companies, oil & gas companies and governments sitting on vast repositories of core samples and geological data hold the key to unlocking new efficiencies and discoveries. However, as data becomes the new gold, issues around intellectual property and national security will become ever-more prominent. Navigating these complexities will require careful planning, strategic partnerships and robust protective measures to ensure that the mining & metals sector realizes the full potential of AI and data analytics.