ARTIFICIAL INTELLIGENCE

“Artificial intelligence (AI) is now an integral part of new chemistry development and is set to supercharge the future of material engineering and reduce the time to discover, test, and deploy new materials and designs.” – Council on Foreign Relations – Leapfrogging China’s Critical Minerals Dominance

This statement from the influential report Leapfrogging China’s Critical Minerals Dominance: How Innovation Can Secure U.S. Supply Chains, published by the Council on Foreign Relations (CFR) and Silverado Policy Accelerator, underscores a pivotal shift in global resource strategy.^1,3,4 Released on 5 February 2026, the report argues that the United States cannot compete with China through conventional mining and processing alone, given Beijing’s decades-long entrenchment across the critical minerals ecosystem-from extraction to magnet manufacturing.^1,2 Instead, it advocates ‘leapfrogging’ via disruptive technologies, with artificial intelligence (AI) positioned as a transformative force in accelerating materials discovery and engineering.^1,4

Context of the Quote and Geopolitical Stakes

Critical minerals-such as rare-earth elements (REEs), lithium, cobalt, and nickel-are indispensable for advanced technologies, including electric vehicles, renewable energy systems, defence equipment, and semiconductors.^1,5 China dominates this sector, controlling over 90% of heavy REE processing and nearly all permanent magnet production, creating strategic chokepoints that it has weaponised through export controls since 2023.¹ In October 2025, Beijing expanded restrictions on REEs and related technologies, nearly halting global supply chains and exposing U.S. vulnerabilities.¹

The report emerges amid escalating U.S.-China tensions under the second Trump administration, where retaliatory tariffs and bans on semiconductor inputs like gallium and germanium have intensified.¹ Traditional responses, such as expanding domestic mining, face insurmountable hurdles: multi-year permitting, billions in upfront costs, environmental concerns, and China’s unmatched scale.^1,2 The quote highlights AI’s potential to bypass these by supercharging chemistry and materials engineering, slashing discovery-to-deployment timelines from decades to years.¹

Authors and Their Expertise

The quote originates from a report co-authored by two leading experts in geoeconomics and supply chain policy.

• Heidi Crebo-Rediker, Senior Fellow for Geoeconomics at CFR and a member of Silverado’s Strategic Council, brings deep experience from her time as U.S. State Department Chief Economist (2014-2017) and roles at Goldman Sachs and the National Economic Council. Her work focuses on financial sanctions, economic statecraft, and resilient supply chains.^3,4
• Mahnaz Khan, Vice President of Policy for Critical Supply Chains at Silverado Policy Accelerator, specialises in frontier technologies and mineral security. Silverado, a non-partisan think tank, drives innovation in national security challenges, and Khan’s contributions emphasise pragmatic financing and allied cooperation to scale breakthroughs.^3,4

Endorsed by CFR’s Shannon O’Neil, Senior Vice President of Studies, the report calls for embedding innovation-including AI-driven materials engineering-into U.S. policy, alongside waste recovery, substitute materials, and international frameworks like the Forum on Resource Geostrategic Engagement (FORGE).^2,4

Leading Theorists in AI-Driven Materials Science and Critical Minerals

The report’s vision aligns with pioneering work at the intersection of AI, chemistry, and materials engineering, where theorists and researchers are revolutionising discovery processes.

• Alán Aspuru-Guzik (University of Toronto) is a trailblazer in AI for molecular discovery. His Molecular Space Exploration Engine (MOSE) and A-Lab-a fully autonomous lab-use reinforcement learning and generative models to design and synthesise novel materials, such as battery electrolytes, in weeks rather than years. Aspuru-Guzik’s ‘materials genome’ approach treats chemical space as a vast data landscape for AI navigation, directly supporting faster REE substitutes and magnet alternatives.¹
• Roald Hoffmann (Nobel Laureate in Chemistry, 1981), though not an AI specialist, laid foundational theories in extended Hückel molecular orbital methods, enabling computational simulations that AI now accelerates. His work on chemical bonding informs AI models predicting material properties under extreme conditions, vital for critical minerals applications.
• Andrea Goldsmith (Stanford) and collaborators in AI-optimised catalysis advance sustainable extraction from tailings and waste-key report recommendations. Their models integrate machine learning with quantum chemistry to design enzymes and photocatalysts for REE recovery, reducing environmental impact.¹
• Jeremy Keith (EPFL) leads in generative AI for inorganic materials, developing models like M3GNet that predict properties across millions of crystal structures. This underpins high-throughput screening for rare-earth-free magnets, addressing China’s heavy REE monopoly.¹

These theorists converge on a paradigm where AI acts as an ‘oracle’ for inverse design: specifying desired properties (e.g., magnet strength without dysprosium) and generating viable compounds. Combined with robotic labs and quantum computing, this could cut development times by 90%, aligning precisely with the report’s leapfrogging imperative.^1,4

Implications for Materials Engineering

AI’s integration promises not just speed but resilience: engineering alloys resilient to supply shocks, recycling magnets from e-waste at scale, and bioleaching minerals from industrial byproducts.¹ U.S. investments, like the $1.4 billion in rare-earth magnet recycling (November 2025), exemplify this shift, targeting firms like MP Materials and ReElement Technologies.¹ By prioritising innovation over replication, the West can forge secure supply chains, diminishing China’s leverage and powering the next industrial era.