There is a quiet crisis embedded in every electric vehicle motor, every wind turbine, and every guided missile system deployed by NATO allies: the permanent magnets that make them work are almost entirely manufactured in China. Roughly 90% of the world’s permanent magnet supply flows through a single country — a concentration of industrial power that keeps defense planners, automakers, and energy executives perpetually on edge. A Munich-based startup called alqem thinks artificial intelligence can dissolve that dependency, and investors are starting to believe it.
alqem announced the close of an 8 million euro pre-seed funding round, co-led by two well-regarded venture firms: UVC Partners, a Munich-based deep tech investor, and Union Square Ventures, the New York firm known for backing transformative technology companies early. The raise positions alqem at the intersection of two of the most consequential trends in technology right now — AI-accelerated scientific discovery and the urgent scramble to rewire global supply chains.
Built by the Team That Mapped the Materials World
What gives alqem immediate credibility is its founding team. The company was built by the researchers behind Alexandria, widely recognized as the world’s leading open materials database. That pedigree matters enormously in a field where the quality of training data determines everything. Materials science is notoriously difficult to model computationally — the search space of possible atomic configurations is effectively infinite, and synthesizing candidates in a lab is expensive and slow.
alqem’s answer to this challenge is a proprietary AI platform that combines a database of computationally predicted materials with high-quality training data covering a broad range of material properties. Rather than relying purely on known compounds catalogued from existing literature, the platform can reason about materials that have never been synthesized — identifying candidates with desirable properties before a single gram of powder is ever pressed or sintered.
The Rare-Earth Problem Is a National Security Problem
The immediate commercial target is stark and well-defined: rare-earth-free permanent magnets. Today’s high-performance magnets — the kind used in EV drivetrains and offshore wind generators — are built around neodymium and dysprosium, elements extracted and refined almost exclusively in China. Geopolitical tension, export controls, or simple policy shifts in Beijing can send shockwaves through industries that took decades to build.
Finding magnets that perform comparably without rare-earth elements has been a materials science holy grail for years. The problem is that traditional discovery methods are glacially slow. Synthesizing, testing, and iterating on candidate compounds through conventional laboratory work can take years per material class. AI changes the economics of that search dramatically — compressing timelines that once stretched across decades into months.
As reported by ChemEurope, alqem’s broader thesis extends well beyond magnets. The platform is designed to identify novel materials across a range of applications where performance and supply chain resilience are both critical considerations.
Why the Timing Is Right
The funding comes at a moment when governments and corporations alike are throwing serious money at supply chain diversification. The U.S. Inflation Reduction Act, the EU’s Critical Raw Materials Act, and a wave of defense procurement reforms all point in the same direction: the era of assuming stable access to Chinese-processed materials is over.
That political urgency creates a commercial window that alqem is well-positioned to exploit. Early-stage materials discovery has historically struggled to attract venture capital because the timelines are long and the path to revenue is unclear. But the combination of AI-accelerated discovery, a high-quality proprietary dataset, and an explicit supply chain narrative gives investors a story they can underwrite.
With UVC Partners and Union Square Ventures backing the initial sprint, alqem now has the runway to move from computational prediction toward real-world validation — the step that will determine whether its AI platform can deliver materials that actually work outside a simulation.
The materials of the future are being designed by algorithms today. If alqem’s bet pays off, the magnets powering tomorrow’s EVs and wind turbines might not have a single gram of Chinese rare earth in them.
