Between 2023 and 2026, Beijing imposed export controls on gallium, rare earths, and at least seven other materials, with those controls also covering many essential components produced with these materials, such as permanent magnets and semiconductor wafers [S2].
Early actions on gallium, germanium, graphite, and antimony tightened supply and pushed prices up by 15-50 percent, implying a long-term price floor that no fixed-price offtake contract can absorb without a reopen clause [S4]. The 7 rare earth elements subject to the April 2025 licensing regime are scandium, yttrium, samarium, gadolinium, terbium, dysprosium, and lutetium, all medium and heavy REEs with magnetic, optical, and catalytic roles in defense, EVs, wind, and industrial motors [S5].
Why a plain offtake no longer covers the risk
A traditional long-term agreement priced in USD per kg and tied to a single mine or refinery assumes the seller can deliver. Under the 2024 dual-use regulations, the Ministry of Commerce can require an export license for Chinese-origin dual-use items already abroad, for foreign-made items that contain designated Chinese-origin dual-use inputs, and for foreign-made items manufactured using designated Chinese technology, meaning the license gate sits downstream of the offtake signature [S4]. The October 9, 2025 control package would have extended licensing to five additional rare earths and obliged any exporter shipping components between third countries to obtain a Chinese license, even if those parts contained only traces of China-produced rare earths; Beijing postponed that package to November 2026, which alone is enough to void the price assumptions in a 5-year contract [S2].
Concentration is the underlying driver: China accounts for around 91 percent of refined rare earth oxide output and 92 percent of global rare earth permanent magnet production, which is monopoly-like leverage that turns any license delay into a force majeure event for the buyer [S3]. For procurement teams specifying control cable insulation compounds or cable gland brass that contains trace antimony or bismuth, the same gate applies, because the licensing trigger is material-based, not end-product-based.
Contract clauses that now have to be written in
Three clause families have moved from optional to mandatory in any 2026 long-term critical-mineral offtake. First, a license pass-through: the seller must disclose whether any covered item, including sub-components such as NdFeB magnets inside an EV traction motor assembly, is on a Chinese control list, and must name the licensing authority and the typical review window, which has run from 30 days to indefinite since 2023 [S1][S2]. Second, a price-revision band tied to a published index: with gallium and germanium moving 15-50 percent inside a single quarter during 2024, fixed prices are commercially uninsurable, and most banks now refuse to back LCs on fixed-price mineral offtakes [S4]. Third, a dual-source or substitution obligation, with defined qualification timelines for the backup supplier, because the IEA has flagged that supply concentration risks have already become reality for several magnet and battery inputs [S1].
For a procurement engineer, the practical drafting test is simple: if the contract cannot name the 7 controlled REEs, the 4 controlled minor metals, and the extraterritorial license trigger in a single schedule, the contract is not paper-ready for 2026 delivery.
Who benefits, who is exposed
Beneficiaries are sellers with non-Chinese refining capacity under binding offtake, mid-stream processors qualified under the G7 Critical Minerals Resilience and Production Alliance, and trading houses that can intermediate license risk. The G7 defines excessive market share as 60 percent or more of a single supplier across the REPM supply chain, which is the threshold used to gate qualified imports under the proposed Economic Security Agreement [S3]. Exposed parties are single-source buyers of NdFeB magnets, gallium-arsenide wafers, and graphite anodes, plus any OEM whose bill of materials contains covered items even at the 0.1 percent trace level, because the October 2025 draft would have captured those shipments [S2]. The asymmetry is intentional: Chinese authorities have signaled that controls are calibrated for leverage, not for permanent withdrawal of supply, which is a game-theoretic posture that keeps the threat live even when volumes partially recover [S5].
For OEMs that route assemblies through cable drag chain systems on automated lines, a 6-week magnet delay can idle an entire cell, so dual-source qualification of the motor and its cable harness is now a contract-term, not a purchasing nice-to-have. Related coverage of this exposure is in EV traction motor magnet supply risk: NdFeB dependence in 2026, which maps the same concentration numbers to specific drivetrain programs.
Limits of supply-side fixes without demand-side rules
State support through grants, equity stakes, and offtake agreements for specific materials is essential but insufficient, because diversified supply is a public good with higher costs attached and individual firms are unwilling to pay the premium alone [S2]. A plurilateral rare earth permanent magnet Economic Security Agreement would convert that voluntary choice into an enforceable market rule, using a capacity-linked tariff-rate quota applied to single suppliers with 60 percent or more market share, with non-qualified imports retaining tariff-free access only up to the volume that qualified supply cannot yet meet [S3]. Until such a rule is in force, the OECD data shows export restrictions still vary widely across countries and material categories, which is exactly the patchwork that fixed-price offtakes cannot hedge [S7].
For a procurement team, the working assumption through 2026 is that any long-term contract on a covered material needs three things in writing: a named license trigger, an indexed price band wide enough to absorb a 50 percent spike, and a qualified backup source with a documented audit trail. Track the November 2026 implementation of the postponed October 9, 2025 control package [S2], the next quarterly OECD export-restriction dashboard [S7], and any G7 Critical Minerals Alliance action on the proposed Economic Security Agreement [S3] as the three signals that will reset contract terms through 2027.