Rare-earth controls are a supplier-exposure ownership test.

Rare-earth controls are not just an upstream story.

On 4 April 2025, China's Ministry of Commerce and General Administration of Customs issued Announcement No.18 of 2025, applying export controls to some medium and heavy rare-earth related items. The official English version lists item families linked to samarium, gadolinium, terbium, dysprosium, lutetium, scandium and yttrium, including certain oxides, compounds, alloys, targets and permanent magnet materials.

The announcement says exporters of the listed items must apply for a licence and identify controlled items in customs declarations. It also says customs can question doubtful declarations and hold goods during the questioning period. That does not make the policy a general export ban. MOFCOM later said the controls are not bans and that eligible applications can receive licences. But for a buyer several tiers downstream, a licensing control can still become a practical evidence problem.

The exposure may not sit where the spend sits.

Many manufacturers track supplier exposure through spend, direct supplier quality and delivery performance. Rare-earth controls cut across that view. The controlled input may sit below the direct supplier. It may be in a magnet powder, alloy, motor, actuator or assembly that appears only as a component line in the buyer's system. The commercial promise, however, is visible: the quote, lead time, service obligation, margin and customer delivery date.

That is why the first useful move is not a critical-minerals lecture. It is a supplier-exposure map. The map asks which products depend on controlled inputs, which supplier node holds the evidence, which documents would be needed if the route tightened, what substitution options are real and which customer commitments would feel the pressure first.

Use a control map, not a concern list.

A concern list says rare earths are important. A control map says who in the business can prove where they matter. For a board discussion, the map should have seven lanes.

• Controlled input. Name the likely rare-earth material or item family, even if the first answer is only "possible dysprosium-containing magnet material" or "supplier confirmation needed".
• Product dependency. Link the input to the product family, component or process that would be affected.
• Supplier node. Identify whether the evidence sits with a direct supplier, component supplier, distributor, exporter, importer or customer-specified component route.
• Licence and documentation friction. Record what the supplier can evidence about licensing, controlled-item status, customs declarations and route ownership.
• Substitution or domestic capability. Separate real alternatives from comforting slogans. A UK programme or recycling route is useful context, not automatic cover for this product line.
• Commercial trigger. Name the event that would make the issue visible, such as quote validity, stock limit, lead-time change, customs delay, customer contract date or margin exposure.
• Evidence owner. Assign a named function or person to close the proof gap.

UK policy is moving in the same direction: visibility, resilience and capability.

The UK is not treating critical minerals as a distant technical topic. GOV.UK's Critical Minerals Programme, last updated on 5 August 2026, describes three pillars: the Magnet Hub, the Critical Minerals Accelerator and the Demand Aggregation Platform. The Magnet Hub is intended to support rare-earth permanent magnet manufacturing, recycling, partnerships and skills. The Demand Aggregation Platform is intended to clarify UK industry demand and support strategic offtake.

That matters because it points to a policy response, not an immediate company-level answer. A future domestic capability may reduce certain dependencies. It does not tell today's manufacturer which supplier, product line or customer promise is exposed. The business still needs its own evidence map.

Supply-chain risk hides in networks, not neat tiers.

The Government Office for Science made the wider point in June 2026 when it published its Foresight work on supply-chain resilience. GOV.UK's summary says supply chains are complex networks, and that a product may look secure while depending on a wider web of interconnected components and materials where vulnerabilities can emerge in unexpected places.

That is the rare-earth problem in board language. If the business only asks its direct supplier whether the next delivery is on time, it may miss the evidence question below the delivery promise. If it asks where controlled inputs sit, who owns the documentation and which commitments would be affected, it has a management control.

A worked example.

A manufacturer quotes a customer for equipment that includes a bought-in electric motor. The motor comes from a European supplier. The supplier has been reliable, the part number is approved and procurement sees no obvious China exposure. But the motor uses permanent magnets, and the magnet material may depend on rare-earth inputs that sit further upstream.

The useful question is not whether the UK manufacturer can redesign the motor by next week. It is whether anyone can answer six questions: which motor lines use relevant magnet material, which supplier can evidence the material route, whether any controlled-item declaration is involved, whether an alternative motor or magnet route exists, which customer commitments would be affected first and who owns the evidence gap.

If the answer is scattered across procurement, engineering, quality, legal and sales, the board does not yet have a supplier-exposure control. It has fragments of one.

Red flags.

• The business says it has no rare-earth exposure because it does not buy raw material.
• The bill of materials names components but not likely controlled input dependencies.
• Supplier assurance sits in procurement emails rather than a governed evidence file.
• Customer delivery promises have no link to supplier evidence quality.
• Substitution is described as possible, but no approved alternative route is named.
• Critical-minerals policy updates are monitored by no named owner.

The decision rule.

Do not treat critical-minerals exposure as solved because the business does not buy raw material. A controlled input can sit three suppliers away while the commercial exposure sits in this quarter's customer promise. The board-ready move is a Supplier Exposure Control Map: product line, supplier node, controlled input, documentation friction, alternative route, commercial trigger and evidence owner. That turns a distant policy story into a manageable operating question.