By KoBold MetalsLearning Objectives
- Understand how AI is applied to mineral exploration
- Understand why critical metals make this a high-stakes application
- Evaluate KoBold's claims and its shift from explorer to mine operator
What Is KoBold Metals?
KoBold Metals is the flagship example of AI in mineral exploration. Finding a new ore deposit is a needle-in-a-haystack problem: the obvious near-surface deposits were found long ago, and what remains is deeper and harder to detect. KoBold applies machine learning to enormous, disparate geoscience datasets — geology, geophysics, geochemistry, and decades of historical drilling records — to predict where undiscovered deposits are most likely to be, focusing on the critical metals that electrification and AI-era power demand require: copper, lithium, cobalt, and nickel. Instead of prospectors' intuition alone, it treats exploration as a data-and-probability problem.
Backed by an unusually prominent investor group — including Breakthrough Energy Ventures and Andreessen Horowitz — at a multi-billion-dollar valuation, KoBold has moved beyond pure exploration toward becoming an operator. Its Mingomba project in Zambia, which the company has described as one of the largest copper discoveries in a century, broke ground in 2026 for a multi-billion-dollar mine. That is the arc worth understanding: an AI-native company using models to find metal, then building the mine itself. The honest framing matters here — the "largest in a century" superlative is a company characterization, and turning AI-guided targets into producing mines still takes years of drilling, permitting, and enormous capital. KoBold's models sharpen where to look; geology and economics still decide what gets built.
💡Key Concept
Exploration as a data problem: KoBold reframes finding deposits from intuition-led prospecting to machine learning over all available geoscience data — predicting the probability of mineralization across a region and prioritizing where to drill.
⚠️Warning
Claims are company statements. "One of the largest copper discoveries in a century" is KoBold's own characterization. AI narrows the search, but a promising target still requires years of drilling, permitting, and capital before it becomes a producing mine — and many targets never do.
✅Tip
Visit KoBold Metals: koboldmetals.com — a private, venture-backed AI exploration company turned mine developer.
Pricing
KoBold is not a software product with published pricing — it is a venture-backed exploration-and-development company that finds and now builds mines, funded by investors rather than selling a tool.
- Machine-learning target generation
- Critical-metals focus
- Proprietary use
- Own projects (e.g. Mingomba)
- Explorer turned operator
- Investor-funded
Core Features
Machine-Learning Target Generation
Predicts where undiscovered deposits are most likely by learning from geology, geophysics, geochemistry, and historical drilling.
Critical-Metals Focus
Targets copper, lithium, cobalt, and nickel — the metals electrification and AI-era power demand most require.
Explorer Turned Operator
Has moved from generating targets to developing its own mines, most notably the Mingomba copper project in Zambia.
Prominent Backing
Funded by Breakthrough Energy Ventures, Andreessen Horowitz, and others at a multi-billion-dollar valuation.
Strengths
- Flagship AI-in-mining story — the marquee example of the category
- Data-driven exploration — machine learning over all geoscience data
- Critical-metals focus — aligned with electrification demand
- From finding to building — an AI explorer now developing a major mine
- Exceptional backing — Breakthrough Energy, a16z, multi-billion valuation
Limitations and Considerations
- Headline claims are company statements — treat superlatives with care
- Long, capital-intensive path — targets to mines take years
- Not a sold product — a company, not a tool you can buy
- Geology and economics decide — models sharpen the search, not the outcome
- Execution risk — becoming an operator is a very different challenge
Best Use Cases
| Use Case | Why KoBold Matters | Caveat |
|---|---|---|
| Understanding AI in exploration | The flagship example of the field | Claims are company statements |
| Critical-metals discovery | ML targeting of copper, lithium, and more | Targets still need drilling |
| AI explorer to operator | Now developing the Mingomba mine | Execution and capital risk |
| Electrification supply story | Focused on metals for the transition | Long path to production |
Key Takeaways
- KoBold Metals is the flagship of AI in mineral exploration, using machine learning to predict undiscovered deposits of critical metals
- It focuses on copper, lithium, cobalt, and nickel — the metals electrification and AI-era power demand require
- Backed by Breakthrough Energy and Andreessen Horowitz, it has moved from explorer to developing its Mingomba copper mine in Zambia
- Its "largest in a century" framing is a company characterization; AI narrows the search but mines still take years, permits, and capital
- Its models sharpen where to look, while geology and economics still decide what actually gets built
