As easily accessible surface mineral deposits become increasingly depleted, the search for deeply buried "hidden minerals" has become a new challenge for the global geological community. In the wave of artificial intelligence (AI), Chinese media pointed out that Chinese and American companies are flocking to Africa, applying AI technology to local mineral exploration. Among them, China's "Shenmai Holdings" and the US "KoBold Metals" have become major players.
According to Yicai, in recent years a number of AI mineral exploration companies have emerged, mainly concentrated in China and the United States, such as the American KoBold Metals, which is backed by investors like Bill Gates and Jeff Bezos; and in China, a new group of AI mining companies such as Shenmai Holdings, an enterprise within the Dreame Technology ecosystem, have appeared. “This is both a result of technological progress and current market demand.”
Shenmai Holdings Chairman Feng Yunduan stated: “The easily explored shallow mineral deposits on the surface are increasingly running out, so geologists have no choice but to venture into deeper, riskier, and more complex hidden regions.” High-risk, extreme geographical environments such as high-altitude, hypoxic plateaus, deserts, and tropical rainforests pose safety risks that traditional surveyors cannot address.
He said the company holds mining rights in several countries including Namibia, Pakistan, Mozambique, Nigeria, and Tanzania, with a total estimated value exceeding 10 billion RMB (5.708 billion MYR). Shenmai Holdings adopts the "Deep Vision Exploration" system, combining multispectral and magnetic sensor drones with autonomous sampling robot dogs, to achieve a fully AI-driven process from sample collection, analysis, to ore vein localization. Compared with traditional prospecting, soil sampling that used to take months can now be reduced to a few weeks, greatly lowering costs and improving accuracy.
KoBold Metals, on the other hand, discovered high-grade copper ore at its Mingomba project in Zambia, having already invested over $200 million (787 million MYR), making it the largest American investment enterprise locally. The core area’s copper content exceeds 5%, much higher than the current copper mine average of 0.6%. Company president Josh Goldman pointed out that AI prospecting can narrow exploration scopes and improve investment returns. KoBold Metals has also expanded its exploration to lithium and nickel mines in Namibia.
Although AI prospecting has begun initial application, Fu Xiaofang, senior engineer and professor at Sichuan Integrated Geological Survey Institute, pointed out that there are still three major bottlenecks: data silos, hardware performance, and model generalization ability. Existing AI models are mostly based on specific geological conditions, and prediction accuracy in complex strata such as deep water or unconventional oil and gas still needs improvement. The report notes that although AI mineral exploration is still in its early stages, its potential in high-risk mining areas, overseas resource integration, and precision operations is gradually emerging, and is set to change the global mineral exploration model.