The News
The race to harness AI technology to supercharge the search for mineral deposits across Africa is picking up pace as investors seek an edge in an increasingly competitive landscape.
Investors, executives, and government officials who spoke to Semafor at the Mining Indaba in Cape Town said the technology offered the promise of exposing the value that lies underground across a continent widely considered to be underexplored.
DR Congo this week signed a five-year agreement with US investment company Atlas Park, which will use proprietary AI software to examine historical data on the country’s mineral deposits and carry out new geological surveys to drive investment decisions.
“Nowhere in the world has more potential for exploration than the DRC,” Kai Han, Atlas Park’s CEO, told Semafor. “We want to make money by investing in exploration… and to do that effectively we need to create a better data environment.”
The CEO of Xcalibur Smart Mapping, which uses aircraft to provide detailed maps of natural resources in 15 African countries — including Benin, DR Congo, Nigeria, Sierra Leone, and Zambia — said his company had ramped up its use of AI in recent years.
Andrés Blanco said the technology was used to synthesize data from a range of sources to help analysts make better assessments about the potential presence of minerals. “AI is opening new business opportunities,” said Blanco. “About 85% of the continent is not explored, and that is difficult if you want to invest.”
Know More
Atlas Park’s agreement with DR Congo is the latest in a number of deals between Kinshasa and US companies and Washington itself, as the African nation looks to benefit from its natural resources. Copper and cobalt — of which the central African country has rich deposits — are in particularly high demand because of their use in AI technologies and clean energy.
Han, who declined to disclose the value of the agreement with Kinshasa, said findings from its AI analysis of historical surveys would be shared with DR Congo’s national geological service. He said this would help the country to benefit from the findings, since it could be used to attract more interest from mining companies.
Blanco also said the deployment of AI technology has the potential to benefit African countries. “We think that it will be a tool for African countries to issue financing,” he said. “Today there are bonds based on oil reserves; there will be financing bonds based on mining potential.”
Step Back
Some estimates suggest Africa holds as much as 30% of known key minerals reserves, but others have argued that Africa’s share of global “critical minerals” production and reserves is much lower. Bright Simons, an analyst at the Accra-based think tank Imani, said it was actually less than 5% in a Semafor column last year.
Aside from disagreements over how to categorize minerals, and what are deemed to be “critical,” a major stumbling block in any assessment of the continent’s mineral wealth is the gaps in knowledge about the continent’s resources. “Africa remains largely underexplored, and known deposits remain significantly under-developed,” found a recent World Bank study.
KoBold Metals, a US exploration company backed by Bill Gates and Jeff Bezos that uses artificial intelligence and machine learning to identify battery metal deposits in Zambia, is the best known such company operating in the continent.
New players — such as Los Angeles-headquartered subsurface modelling company Terra AI, and Arizona-based automated drilling firm Durin — are building businesses based on improving mining efficiency. Executives from both companies were in Cape Town for Mining Indaba.
Alexis’s view
AI-driven technology is already changing African mining, and the pace of change looks set to quicken. It’s easy to focus on its use in exploration but that would be overly simplistic. In reality, the technology is a toolkit that enables geologists to draw on data, models, and geophysics to generate valuable insights at various stages of a mining project.
The most obvious impact is the provision of more certainty to attract investment. Demand for the minerals needed for everything from smartphones to computer chips and batteries is sky high, but mining has always struggled with uncertainty about the true nature of underground discoveries. Factor in the often overblown fears of political risk in African countries, and it’s clear that these tools can help to reassure investors.
The danger for African countries is that advanced economies, racing to build their economies for the coming decades, could use AI-driven mining to more efficiently extract valuable minerals without the host countries benefiting. African policymakers will need to design and implement policies that ensure local processing and refining of minerals is factored into their plans. Without that, the use of AI could go down as the point at which the 21st century’s scramble for Africa went into overdrive.
Room for Disagreement
Increased use of AI in mining is set to heighten data security risks due to the vast amount of information to be processed, a research paper by S&P Global warned.
The authors also said the findings produced by AI are only as good as the data used, which could be flawed. The “recommendations from AI systems may yield discriminatory results if there is inherent bias from data sources and existing algorithms,” they said.
The View From Washington
British Robinson, Africa chair at the Washington-based Milken Institute International think tank, said it was important to broaden the definition of “innovation” in conversations about the demand for Africa’s mineral resources.
“Innovation means we need more than one partner,” said Robinson, stressing the need for a radical rethinking of previous approaches to mining in the continent. “That means more partners, different partners, different vehicles, funds that can be deployed for different parts of project phases,” she said, adding that this was needed to ensure that projects are “actually bankable.”
Robinson said the inclusion of a price floor in the US government’s planned critical minerals trade zone was also innovative, citing it as a new approach that would “make investors feel safe.”
Notable
- A framework for digital mineral information rights is needed to create a tradeable asset class from geological intelligence, argued analyst Bright Simons in a recent ODI Global paper.
- AI is likely to change exploration, mining, and the processing of critical minerals in “profound ways,” wrote Jef Caers, an earth and planetary sciences professor, in Global Mining Review.


