Europe's other sovereignty problem: Critical Minerals

Europe has spent the last year worrying about sovereignty. The loudest version of this is compute. When Washington ordered Anthropic to switch its 2 best AI models off for every non-American overnight,¹ the "kill switch"² suddenly had a date on it. Europe 2031, the scenario Judith Dada and seven co-authors published in June on the continent sliding into AI dependence, hit the same nerve.³ Defence is the other headline. Underneath all of it runs a dependency that gets far less attention: the physical inputs. The batteries, magnets, motors and chips behind both the energy transition and the AI boom trace back to a few dozen minerals that Europe barely produces. For 19 of the 20 minerals the IEA classes as strategic, the top refiner is one country, China.⁴

Download the full visual explainer here.

This dependency is neither hypothetical nor free. When China restricted rare-earth magnets in April 2025, the ones inside most EV motors and most offshore wind turbines,⁵ production lines across Europe shut down, and by June Europe's automotive suppliers reported that only about a quarter of their export-licence applications had come back approved.⁶ And demand is only climbing: an EV needs 6 times the mineral input of a combustion car,⁷ and appetite for these minerals rises steeply through the 2030s.⁴ The catch is time: a new mine takes over 16 years on average from discovery to first ore production,⁷ and Europe cannot dig its way out on that schedule.

We can solve this, and it isn't by digging

For the last 2 decades Europe has tried to dig its way out, and the record is thin. Rio Tinto committed 2.4 billion dollars to the Jadar mine in Serbia, once slated as Europe's largest lithium source. Permits were revoked in 2022 and reinstated in 2024, but Rio itself still shelved the project in November 2025.⁸ Finland's Keliber did get there and began mining this February, 17% over budget at about 880 million dollars.⁹

None of this says Europe should stop building. It says new mines are slow, expensive and few, and cannot be the whole answer on the timeline that matters. The better starting point is that much of what Europe needs has already been dug once: it sits in scrap, tailings and spent batteries. Supply is gated by 2 things: speed and access. Technology can move both faster and closer to home than a new pit. Three levers follow, in rough order of how fast each can move.

“The cheapest mine is the one Europe never digs.”

Lever 1: Reduce the need for new rock

Recycling is the most commercially mature of the 3 levers, and the reason is feedstock: factory scrap sits inside Europe today, dead battery packs arrive in volume through the 2030s from the current generation of EVs, and little of either is in the mining industry's hands. The EU Battery Regulation will demand recovery rates of 90% for a spent battery's nickel and cobalt from the end of 2027, lithium on a rising ladder behind it.¹⁰ Cylib in Aachen and tozero in Munich are scaling it now.¹¹ Recupere turns 100% recycled scrap into electrical-grade copper wire and says its process removes the need for smelting entirely.¹² RarEarth in Italy goes after the harder target: recovering rare-earth magnets from end-of-life motors.¹³

The less visible end of this lever is designing the constrained input out of the supply chain entirely: UP Catalyst turns captured CO₂ into battery-grade graphite,¹⁴ Altrove uses AI to find material chemistries that reduce or remove the scarce input,¹⁵ and sodium can stand in for lithium where the application allows.⁴

Lever 2: Get more from what's already being dug

Every operating mine on earth is already through discovery, permitting and construction, and the extra yield comes from better processing, plus the tailings the first pass left behind.⁴ Biomining and biorefining swap the smelter for engineered microbes and proteins that pull metal from low-grade ore and waste. Vienna's Magmatic Bio designs synthetic proteins that bind and selectively extract metals during refining.¹⁶ France's Genomines uses genetically enhanced plants to recover nickel from soils too poor to mine conventionally.¹⁷ Colorado's Endolith engineers microbes that pull more copper from ore already being leached, lifting yield at mines that cleared permitting long ago.¹⁸

Lever 3: Fast-track the few mines we genuinely need

The first 2 levers cut how many new mines Europe needs; this one makes the unavoidable ones faster. Paris-based Lithosquare builds geology AI that narrows where to drill for copper, lithium and rare earths, cutting months of analysis to days.¹⁹ Direct lithium extraction pulls lithium straight out of brine with sorbents and membranes instead of open pits and evaporation ponds: less land equals a shorter permit file. Vulcan Energy is the standout European case: its plants pump hot brine from deep under the Upper Rhine to make geothermal power, filter out the dissolved lithium, and send the water back underground. It closed a 2.2 billion euro financing package in December 2025, fully funding its first phase, and took the first lithium production licence of its kind in the Upper Rhine Valley Brine Field a little over 3 months later.²⁰

The real race is time

By now the shape of the answer is clear, but time is of the essence. The demand curve everyone can see coming bends up in the early 2030s, and a deposit found today, on a 16-year clock, delivers ore well after that curve turns. Even recycling, the lever we like best, arrives in stages: production scrap and end-of-life copper can be processed today, while the lithium and nickel locked in EV packs only reach recyclers in volume in the 2030s.⁴

In Brussels, the people paid to see this coming have started to move. The EU's Critical Raw Materials Act has designated 47 strategic projects inside the EU and 13 beyond it,²¹ and the RESourceEU plan, announced in December, promises up to 3 billion euros across 2026.²² That tells you where the next decade of industrial policy, and demand, is heading. The money is moving more slowly: the Commission says it has mobilised 1.7 billion euros since December, and 23 of the 60 designated projects signed a letter to Brussels in August warning of liquidity and financing pressure.²³

Every tonne recovered closer to home is supply that lands this decade, and control that doesn't sit in Beijing. On the IEA's 2024 numbers the market was already worth around 325 billion dollars a year, and the agency sees it more than doubling to 770 billion by 2040 in its net-zero scenario.²⁴

We spend our time and our capital here. If you're building in recycling, alternative materials, smarter extraction, or faster routes from discovery to production, I'd like to hear from you. And if you think we've got a piece of this wrong, curious to hear that too.

Sources

1. Fable 5 / Mythos 5 export-control directive (12 June 2026); scope = any foreign national anywhere, incl. Anthropic's own foreign-national staff; models disabled worldwide: Anthropic statement, 12 Jun 2026 (para 1: "suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees"; "we must abruptly disable… for all our customers") · CNBC, 12 Jun 2026 · Fortune, 13 Jun 2026 (names the Commerce Department)
2. EU Technological Sovereignty Package (3 June 2026); "kill switch" quote (Executive Vice-President Henna Virkkunen, on cloud providers of critical workloads): European Commission news page, 3 Jun 2026 · CNBC, 3 Jun 2026 (quote in headline; attribution in body)
3. Europe 2031 scenario, published 11 June 2026 by Daan Juijn and Judith Dada with 6 co-authors: europe2031.ai · About · Launch post. Dada: GP, Visionaries Club; Juijn: Director of Research, Arq Foundation.
4. IEA, Global Critical Minerals Outlook 2025, page-pinned per claim: 19-of-20 minerals refined by China, ~70% avg (exec summary; PDF p.8) · >95% of battery-grade graphite (PDF p.30) · 91% of refined rare-earth output (PDF p.175) · top-3 refining nations 82% (2020) → 86% (2024) (exec summary; PDF p.6) · lithium demand 5× by 2040, graphite and nickel 2×, STEPS (PDF p.89) · battery-cell production cost 40–50% higher in Europe/US than China (PDF p.98) · sodium-ion as upstream diversification (PDF p.10) · tailings re-mining as a supply lever (PDF p.9) · secondary supply ramps post-2030 as end-of-life feedstock arrives (PDF p.99). Exec summary · Full PDF
5. 7 heavy rare earths + NdFeB magnets (4 Apr 2025); 0.1% extraterritorial magnet rule (9 Oct 2025): EPRS, "China's dominance in rare earth elements" (PE 779.220, Nov 2025), p.1 · US-specific general licences under the November deal: White House fact sheet, Nov 2025 ("China will suspend the global implementation… will issue general licenses… for the benefit of U.S. end users") · April controls in force and actively enforced as of Jul 2026: Morgan Lewis client alert
6. CLEPA: ~25% of export-licence applications approved; European supplier lines shut down; BMW said part of its supplier network was affected: CLEPA statement, 4 Jun 2025 ("only approximately one-quarter appear to have been approved") · CNBC, 5 Jun 2025 · Reuters factbox via Yahoo Finance, 5 Jun 2025
7. EV ≈ 6× the mineral input of a conventional car; onshore wind ≈ 9× a gas plant; 16.5-year average from discovery to first production: IEA, The Role of Critical Minerals in Clean Energy Transitions (2021), exec summary
8. Jadar: 2004 discovery; permits revoked Jan 2022; reinstated Jul 2024; shelved by Rio Nov 2025 (care and maintenance); $2.4bn committed (2021); near-$3bn project: MINING.COM, 13 Nov 2025 · electrive, 17 Nov 2025 ("largest lithium reserves in Europe") · RFE/RL, 16 Jul 2024 (reinstatement) · Rio Tinto, 27 Jul 2021 ($2.4bn commitment)
9. Keliber (Sibanye-Stillwater): Syväjärvi mine began operations 11 Feb 2026; capex raised 17% to €783M (~$880M): electrive, 12 Feb 2026 · MINING.COM, 9 May 2025
10. EU Battery Regulation 2023/1542 recovery targets: by end-2027, 90% cobalt/copper/lead/nickel and 50% lithium; by end-2031, 95% and 80%: European Commission
11. cylib (Aachen) and tozero (Munich), battery recycling at industrial scale: cylib.de · tozero.solutions · tech.eu, 27 Mar 2026 (tozero industrial plant, >1,500 t/yr)
12. Recupere: copper wire from 100% recycled feedstock; "our process removes the need for smelting"; €5M seed (SISTAFUND): recupere-metals.com · Funding announcement
13. RarEarth (Italy): NdFeB magnet recovery from end-of-life motors and electronics; €2.6M round led by Primo Capital: rarearth.it (use the www host; the apex domain's TLS is broken) · i3p announcement
14. UP Catalyst: CO2-derived battery-grade graphite (MSCC-ET process); €18M EIB financing: upcatalyst.com/about-us · EIB financing announcement
15. Altrove: AI-driven materials discovery to reduce or remove rare-earth dependency: altrove.ai · Alven investor announcement ("rare-earth-free… formulations")
16. Magmatic Bio (Vienna): custom synthetic proteins that bind and selectively extract metals during (bio)refining: magmatic.bio
17. Genomines (France): genetically enhanced plants recovering nickel from soils "too low-grade to mine traditionally, but too metal rich to farm"; $45M Series A: genomines.com · renewable-carbon.eu
18. Endolith (near Denver): engineered microbes enhancing copper recovery from existing heap leach: endolithmining.com/technology · Scientific American
19. Lithosquare (Paris): geology AI for copper, lithium and rare earths; analysis timelines "from months to days"; $25M round: Omnes Capital press release · World Fund portfolio page
20. Vulcan Energy: €2.2bn financing (Dec 2025) fully funding Phase One (Lionheart); first lithium production licence issued in the Upper Rhine Valley Brine Field and in Rhineland-Palatinate (17 Mar 2026); geothermal brine mechanism: electrive, 3 Dec 2025 · Mining Weekly, 17 Mar 2026 · EIB (mechanism; "largest lithium resource in Europe")
21. CRMA: 47 strategic projects designated within the EU (first round, Commission Decision (EU) 2025/840, 25 Mar 2025) and 13 in third countries (Jun 2025): European Commission, 19 Jan 2026
22. RESourceEU: "Up to €3-billion will be mobilised over the next year" (announced Dec 2025): Mining Weekly, 4 Dec 2025
23. 23 of the 60 designated CRMA strategic projects signed an "Urgent Call to Action" to the Commission over liquidity, financing and market access (Sep 2026): Reuters, 8 Sep 2026 · EU Today, 9 Sep 2026
24. Combined market value of key energy transition minerals ≈ USD 325bn today, more than doubling to USD 770bn by 2040 in the NZE Scenario: IEA, Global Critical Minerals Outlook 2024, exec summary ("At around USD 325 billion, today's aggregate market value of key energy transition minerals aligns broadly with that of iron ore… more than doubles to reach USD 770 billion by 2040 in the NZE Scenario")