發生了什麼事
KoBold Metals, a California‑based junior mining company founded in 2018 with investors that include Bill Gates and Jeff Bezos, has begun field operations in the Democratic Republic of Congo (DRC) using artificial‑intelligence tools to locate lithium deposits. Geologists in the field carry GPS‑enabled tablets that connect in real time to an AI system hosted on KoBold’s servers. The AI ingests historic geological maps, legacy drill data dating back to the colonial era, and satellite imagery to produce predictive models that highlight “geochemical signatures” of lithium‑bearing pegmatite. The company says the system has already helped it prioritize drilling locations, allowing teams to sample roughly 2,000 soil cores per week. Since 2025 KoBold has obtained 14 exploration permits covering more than 3,000 km² around Manono in the former Katanga province. The firm plans to invest over $50 million in the DRC by early 2027 and hopes to move from exploration to production within a few years. KoBold’s CEO in the DRC, Benjamin Katabuka, told AFP that results are “very encouraging,” though no specific resource estimates were disclosed.
KoBold Metals has set up a base camp in Manono, a remote town in the mineral‑rich Katanga region, and equipped its field teams with high‑ GPS devices and tablets linked to an AI platform. The AI continuously processes incoming soil sample data, satellite images, and historic drilling records to generate predictive maps that indicate where lithium‑bearing pegmatite is most likely to occur.
Since 2025 the company has secured 14 exploration licences covering more than 3,000 km². The licences were obtained after a strategic partnership agreement between the United States and the DRC government was signed in December of the previous year. KoBold’s leadership says the AI system has already helped prioritize drilling locations, allowing the team to collect roughly 2,000 soil samples each week.
KoBold plans to invest over $50 million in the DRC by early 2027, with the goal of moving from exploration to production within a few years. The company has not disclosed any formal resource estimates, and the timeline for a commercial mine remains uncertain.
為什麼這很重要
The deployment of AI in mineral exploration marks a shift toward data‑driven resource discovery, potentially accelerating the identification of critical minerals needed for electric‑vehicle batteries and renewable‑energy technologies. By leveraging AI, KoBold aims to compete with Chinese mining giants such as Zijin, which currently dominate the DRC’s copper and cobalt sectors and now also produce lithium concentrate. If successful, the project could diversify global lithium supply chains, reduce reliance on China, and give the United States a strategic foothold in a region that supplies a large share of the world’s critical minerals. The initiative also illustrates how high‑profile investors are channeling capital into AI‑enabled mining, signaling broader industry interest in applying advanced analytics to traditionally labor‑intensive sectors. However, the venture operates in a region plagued by corruption, weak governance, and armed militia activity, raising questions about the social and environmental impacts of accelerated mining activity.
AI‑driven exploration could dramatically shorten the time needed to locate viable lithium deposits, a critical input for electric‑vehicle batteries. Faster discovery may help meet the International Energy Agency’s projection that global lithium demand will more than double by 2030.
The project challenges the near‑monopoly that Chinese firms hold over the DRC’s critical‑mineral supply chain. By introducing a U.S.‑backed, AI‑enabled competitor, the venture could encourage more competitive pricing and improve labor and environmental standards if the presence of a new player forces incumbents to respond.
High‑profile investors such as Gates and Bezos signal confidence in the commercial viability of AI‑enhanced mining, potentially spurring further private‑sector capital into similar projects worldwide.
互動機制：它實際上是如何運作的
以互動方式探索這項發展背後的基礎技術。
Which component of an AI application is the machine-learning model itself?
接下來看什麼
Key indicators to monitor include: (1) any public disclosure of resource estimates or drill results generated by the AI system; (2) the timeline and scale of KoBold’s transition from exploration to pilot production; (3) additional U.S. or private investment flowing into the DRC mining sector in response to KoBold’s activities; (4) reactions from Chinese mining firms and any policy shifts by the Congolese government; and (5) community and environmental responses, especially regarding labor conditions and water usage in the Manono area.
Public release of AI‑generated resource estimates or drill results, which would validate the technology’s effectiveness.
Progress on securing additional financing or partnerships, especially from U.S. government programs aimed at securing critical minerals.
Policy responses from the DRC government, including any changes to mining regulations or security measures in the Manono region.
Reactions from Chinese mining companies, which may adjust their own exploration strategies or lobby for protective measures.
Local community feedback, particularly regarding employment conditions, water usage, and environmental safeguards.