Last Updated: 3 June 2026
Terra AI, a startup that uses artificial intelligence to help mining and energy companies map what lies below the earth's surface, has raised $20 million in a Series A funding round. Khosla Ventures led the investment, with BHP Ventures joining as a strategic backer. The company was co-founded by John Mern, who serves as CEO, and Anthony Corso.
Finding and developing underground resources has become a serious bottleneck for the global energy transition. According to the IEA's Global Critical Minerals Outlook 2025, copper faces a potential 30% supply shortfall by 2035, the result of falling ore grades, higher project costs, and fewer new discoveries. Lithium is on a similar path, with demand expected to outpace supply by the 2030s. Making matters harder, the average mineral discovery still takes 17 years to reach production. That timeline has not kept pace with how quickly the world now needs these materials.
Khosla Ventures also led Terra AI's $3.4 million seed round in 2023. BHP Ventures is a new addition as a strategic investor. The broader group of backers includes Rio Tinto, Rio Tinto Founders Factory, Storyhouse Ventures, Plug and Play, The TomKat Center for Sustainability, and Climate Capital.
The fundamental problem in mineral and energy exploration is that no one can see underground. Teams collect drilling results, geophysical surveys, and geochemical samples, but the tools they use to analyse each data type have traditionally been separate. That fragmentation slows interpretation and leaves large gaps in what teams actually know about a deposit before they commit to drilling.
The costs of that uncertainty are significant. Mining companies face pressure to bring copper, lithium, rare earth elements, and other materials to market faster, driven by demand from electric vehicles, power grids, AI infrastructure, and battery storage. Carbon storage sites carry the same challenge: choosing a safe, permanent location underground requires a clear picture of the geology. Enhanced geothermal energy projects depend on the same kind of subsurface knowledge. Across all three sectors, exploration uncertainty drives up costs, extends timelines, and increases the risk of committing capital to the wrong location.
"Successful exploration depends on reducing geological uncertainty and progressing opportunities efficiently. Terra AI's technology has the potential to improve drill targeting, support faster prospect evaluation, and help explorers focus capital on the highest-value opportunities."
Laurel Buckner, VP of Ventures at BHP
The new funding will support three areas: building out the company's generative geological modeling engine, deploying the platform at larger enterprise customers, and advancing its subsurface inference technology. That last term refers to the platform's ability to reason about what likely exists underground based on available data. The company is also developing a geology reasoning agent, a software tool that works with operators to plan where and how to explore.
Terra AI intends to move into carbon storage and enhanced geothermal in addition to its existing mining and reservoir work. Both are areas where understanding subsurface conditions is just as important as it is in mineral exploration, and where demand is growing. Data centers, for instance, are increasingly looking at geothermal energy as a source of reliable, low-carbon power. The company's position is that the same core technology applies across all of them.
"This investment enables us to move to the next phase of Terra AI's growth by scaling our generative modeling engine, accelerating enterprise-grade deployments, and advancing the subsurface inference technology the industry needs to meet critical mineral demand."
John Mern, CEO Terra AI
Terra AI was founded by John Mern and Anthony Corso. Its platform takes in the full range of data that geologists collect, including drill core samples, geophysical surveys, geochemical readings, and newer methods like muon tomography, which uses subatomic particles to image dense materials underground. Rather than producing a single interpretation of what might be there, the system generates millions of possible geological models at once. Each one represents a plausible version of the subsurface, and together they map out where confidence is high and where uncertainty remains.
That shift, from a single map to a range of possibilities, changes how teams make decisions. Instead of committing a drilling program based on one interpretation, teams can see which areas carry the most uncertainty and target those specifically. According to Terra AI, commercial engagements have produced 50 to 60 percent reductions in required drilling meters, a 2 to 3x improvement in resource model accuracy, and acceleration of Tier-1 projects by four to five years.
The company draws a comparison to AlphaFold, the AI system that transformed protein structure prediction in biology, as a reference point for the kind of compression in time and cost it is pursuing.
The platform is designed to work across different project types without modification, covering copper, gold, rare earths, reservoir projects, and carbon storage. Current customers include BHP and Rio Tinto, as well as mid-sized and junior exploration companies.
Khosla Ventures has backed Terra AI since its earliest stage, having led the 2023 seed round before returning to lead this Series A. The firm has previously invested in OpenAI, Instacart, Affirm, DoorDash, and Block. BHP Ventures, the venture arm of BHP Group, comes in both as an investor and as a customer, with BHP already using the platform commercially. Rio Tinto participated in the seed round through Rio Tinto Founders Factory and remains part of the investor group.
"The world's ability to discover and develop critical resources is fundamentally constrained by outdated, fragmented exploration. Terra AI is defining a new, AI-native exploration approach with their continuously improving 3D Earth model. We believe their technology will dramatically accelerate the development of the critical minerals the world needs, and we're excited to help them scale this vision globally."
Rajesh Swaminathan, Partner at Khosla Ventures
