Canadian AI is helping the critical mineral industry get metals out of the ground—and into EVs.
Vancouver-based Vrify Technology just raised $12.5 million in a series B round, led by New York VC firm LGVP. The company, which uses artificial intelligence to help geologists discover mineral deposits, says it is already breaking even. But it’s expanding its business as the mining sector approaches its “ChatGPT moment,” said CEO Steve de Jong in an interview.
Investors and companies in a range of fields are hoping 2025 is the year AI begins transforming humankind’s relationship with the physical world, and mining is no exception. KoBold Metals, which is backed by Bill Gates, Jeff Bezos, and Jack Ma and also uses AI for mineral discovery, closed a US$537 million series C round in January.
Vrify’s connection to things we can see and touch is one of the factors that drew LGVP’s Itai Tsiddon. Much of the hype around AI has centred on YCombinator companies “selling to other YC companies,” Tsiddon noted in an interview, whereas companies like Vrify are driving changes in the “real world.”
Over the past few years, AI research hubs in Canada like the Creative Destruction Lab, the Vector Institute and Mila have started working with mining companies to apply AI. Their interest speaks to the edge the technology can give companies in a viciously competitive sector, where entrepreneurs are under constant pressure to lower costs and speed up production. It also reflects the country’s singular status as a place where mining financiers and AI luminaries co-exist at a critical mass. “There’s no place better set up,” said de Jong.
The opportunities are easier to see than the minerals themselves. EV batteries and electrical grid upgrades are expected to drive a sixfold increase in demand for critical minerals like copper, cobalt and nickel by 2040. But companies combing the earth for them face the twin pressures of money and punishing timelines. It takes nearly 16 years on average to take a mine from discovery to production. Yet funding comes and goes with the volatile ups and downs of commodity prices.
The hope is that AI will help small exploration companies narrow down high-value deposits and give investors information that would be costly to obtain by manual drilling alone.
Christopher Pennimpede, CEO of Canterra Minerals Corporation, is working with Vrify on a Newfoundland critical minerals and gold project where prospecting dates back to the 1920s. He’s hoping a century of well-preserved data will be a treasure trove to feed the AI algorithm, which will help the company pinpoint rich deposits.
Over decades, de Jong said, existing mines have uncovered critical minerals that were previously considered waste but are now hot commodities for use in microchips or high-tech motors. Tools like Vrify might be able to supercharge that work, searching historical data from, say, a gold mine for traces of gallium that could be harvested without setting up new mining sites.
AI models have also opened geologists’ minds to “biases” they may unwittingly hold, Pennimpede said. Canterra’s target area in Newfoundland is a case in point. Previous attempts to mine it focused on the northern part of the property, for which the company had much more historical data showing potential deposits. But the AI model identified a few promising areas on the southern end, prompting Pennimpede to dive into the data. There, he found that prospectors in the 1920s had arbitrarily chosen to start their search on the north end of the claim, stopping to mine as soon as they struck pay dirt and leaving other areas untapped.
The result? A new opportunity for Canterra to get critical minerals to market. Coming soon, perhaps, to a battery near you.
Read Shift—The Logic’s authoritative weekly newsletter on automotive technology industry news—for more; and if you know someone who should be reading it, they can sign up here.
