Several startups have announced in the last year that they use AI technologies to detect areas where essential mineral deposits are found. How artificial intelligence contributes to this “chase” according to useful and valuable minerals and what are its great strengths when used for geological prospecting?
How do technology discover deposits that can change the world's resource map
The batteries of electric cars, wind turbines and advanced armament systems around the world are built using huge amounts of raw materials such as cobalt, copper, lithium, nickel and other rare elements. The great stake is to find and exploit these subjects, and a number of methods made available by new technologies seem to change the situation.
It is also a stake of geopolitics, in a world in which China has huge resources of essential minerals, and the US wants to reduce its dependence on the Beijing mineral supply chain and is on the list and controversial for a mineral agreement in Ukraine.
In the summer of 2024, a startup called Kobold announced that he had discovered in Zambia one of the largest copper deposits found anywhere in the world in the last decade.
Kobold Metall is famous for this project in the Copperbelt region (Zambia), an area famous for its copper reserves. In 2022, the company estimated that the warehouse would contain 247 million tonnes of ore, with an average concentration of 3.6% copper, considered of high quality.
In February 2024, the head of the company announced that Mingomba could become “one of the richest underground in the world”, having ores with a content of about 5% copper, is written in an article published by Foreign Polycy.
This discovery was also greeted by the president of Zambia, Hakinde Hichilema, who said that Mina could become one of the largest in the world. The company has over 60 projects on four continents and claims that it has invested about $ 100 million in research and development in 2023.
A typical mining exploration project means analyzing a huge volume of data, and the AI exactly is very good, so many of the activities related to geochemistry, mineralogy, geophysics and remote sensing can be accelerated.
Earthai and the promise of efficiency
Recently, another startup, called Earth AI, announced, quoted by Techcrunch, the detection of promising deposits of critical minerals in areas of Australia that other mining companies have ignored for decades.
Earth has identified copper, cobalt and gold deposits in the Australian province of Northern Territory, as well as silver, molybdenum and tin in another area of New South Wales, also in Australia. Here is the company's statement.
Earth's algorithms are trained to scan quickly and efficiently vast surfaces to find deposits that would otherwise have been ignored, say the company.
Interestingly, Earth has decided to develop their own drilling equipment to prove that the identified locations are as promising as indicating its software. In the latest financing round, Earth has obtained $ 20 million in January 2025.
Not only smaller companies test technologies in mining, but also giants in the field, such as Rio Tinto.
How AI helps in chase after rare minerals
Very brief, the AI can speed up the process of discovering critical mineral resources, reducing costs, but also impact on the environment.
Artificial intelligence is used more and more to discover essential mineral deposits, and a big plus is seen in the analysis of geological and geophysical data. The AI can identify models that indicate the presence of critical minerals, and automatic learning algorithms can detect magnetic, gravitational or electromagnetic abnormalities associated with mineral deposits.
Moreover, AI models can compare data from known areas that they have mineral reserves and can estimate where new deposits could exist. These so-called “predictive maps” can direct geologists to more promising areas.
The AI is still good at something: it can analyze satellite images and data collected by drones to identify certain geological features associated with critical minerals. It is also important that advanced algorithms can create 3D models of the underground structure, to estimate with greater precision the size and composition of a deposit. This reduces unnecessary drilling and minimizes the waste of resources.
The Economist publication offers some examples, in an article about the future of mining. Freeport-McMoran, a copper specialized mining company, uses sensors mounted on trucks, excavators and machines, to collect real-time data, but not only about the quality of the extracted ore, but also about the speed of operations and the performance of the equipment. This information is then entered in AI models that guide the company's equipment.
Mining companies are not the only ones that intensify their prospecting efforts through technology. The United States Geological Service (USGS) and the US Defense Research Agency (DARPA) have launched a project for developing AI tools to collect and evaluate the quality of critical mineral data.
The plan includes the extraction of geospatial data from a vast base of sources, including 100,000 old maps and geological studies, but also the mapping of minerals through georeferencing.
All these progress accelerated the geological mapping processes, and without these new technologies it would have lasted many years and would have been largely manually performed.
After collecting, integrating and organizing various geostitic data sets, the researchers apply “Machine Learning” techniques to identify patterns, make predictions and estimate where it is likely to be minerals in the US soils.
Photo source: dreamstime.com
