How China's Battery Technology Export Controls Affect Global Buyers

This is where the whole story turns, and where the reporting has been thinnest relative to the importance of what happened.

MOFCOM placed LFP cathode technology on two separate regulatory lists in two separate moves, three months apart. The July restrictions set a compaction density threshold of 2.58 g/cm³, capturing fourth-generation LFP. CATL, BYD, maybe three or four other companies produce at this tier. The restriction went on the Catalogue of Prohibited and Restricted Technologies, governed by the Foreign Trade Law.

The October move dropped the threshold to 2.5 g/cm³. Third-generation LFP. This is the mainstream chemistry. Most Chinese EVs. Most storage systems shipping globally. And October's restriction went on a different legal instrument entirely, the Export Control List, which routes through the Central Military Commission.

The difference between these two legal tracks matters more for buyers than the compaction density numbers do, and the compaction density numbers have gotten all the coverage. The Catalogue governs trade. An application processed under the Catalogue gets evaluated on commercial grounds by trade officials whose institutional orientation is toward facilitating exports, because exports generate revenue and economic activity. The Export Control List governs security. An application there gets evaluated by people whose job is threat assessment, whose institutional incentive runs the opposite direction, toward caution, toward denial. A buyer negotiating a cathode technology license does not get to choose which track the application follows. A buyer may not even know which track it landed on until deep into the process, at which point the negotiating dynamics, the timeline expectations, the probability of approval, have all shifted in ways the buyer did not anticipate when the deal was structured.

Between the January draft and the July final text, every parameter revision widened scope. Compaction pressure testing from 300 MPa to 220 MPa. Tap density for iron phosphate from 2.1 g/cc to 1.2 g/cc. Nobody on the outside can say exactly why the final numbers landed where they did, but the direction of every revision is unambiguous.

The 300 Wh/kg threshold on cell and pack exports is a separate mechanism. It sits above commodity LFP at 160-180 Wh/kg, above most mass-production NMC at 230-270 Wh/kg, and catches the high-nickel and silicon-anode cells going into premium platforms for 2027-2028 model years. Mainstream chemistries are climbing toward it. In two or three product cycles it will catch what is currently midrange.

This section is going to be longer than it probably should be relative to the rest of the article, because the equipment situation is the part of the export control story that most affects buyer behavior in the near term and that has received the least adequate treatment.

Wuxi Lead Intelligent Equipment is the world's largest supplier of battery production machinery. Its customers include CATL, BYD, SK On, Panasonic, LG Energy Solution, Samsung SDI, and ACC. Decision No. 58 put winding machines, lamination machines, and liquid injection machines on the dual-use list.

Why does Wuxi Lead dominate? Iteration volume. China built hundreds of gigafactories over the past decade. Each one was a deployment cycle for equipment suppliers. Each deployment generated feedback on yield rates, throughput limits, maintenance intervals, software integration quirks, operator training requirements. Japanese equipment makers, Toray Engineering and Hirano Tecseed among them, served dozens of deployments. European and American ones served fewer. The performance gap between equipment refined over hundreds of deployments and equipment refined over dozens does not appear on spec sheets. Spec sheets can look comparable. The gap appears when the machines are installed and running on a live production floor. First-pass yield. Uptime over six months. How long it takes a new line to hit nameplate capacity. These are numbers that a procurement team evaluating equipment bids cannot see until after the purchase order is signed.

Chinese equipment suppliers built deep integration between their machines and the manufacturing execution systems (MES) that Chinese cell makers use. A Wuxi Lead winding machine does not just wind electrodes. It feeds granular process data into a software stack that tracks every parameter of every cell produced on that machine. When a quality issue surfaces three months later during field warranty claims, the MES data lets the quality team trace the defective cells back to specific machine settings on specific production shifts. Switch to a non-Chinese winding machine and you have two options: rebuild that entire data pipeline at significant cost and delay, or run your production line with a monitoring gap that becomes painful the first time you need to do root-cause analysis on a batch defect. This switching cost is invisible at the purchase decision stage. It is very visible at the production stage. Procurement teams evaluating a transition away from Chinese equipment are, in most cases, not aware of it until their manufacturing engineers tell them, at which point the engineers' assessment is usually that the switching cost is higher than the procurement team assumed.

The parallel I keep coming back to with the equipment controls is not semiconductors, which has been so overused as a comparison in this space that it now functions mostly as a signal that the person writing has not thought about the question very carefully. The parallel is the Toshiba-Kongsberg affair from the 1980s, where Toshiba Machine sold precision milling equipment to the Soviet Navy that enabled quieter submarine propellers and triggered a major export control overhaul. That case was specifically about machines, about how controlling the equipment that makes things provides leverage over capability rather than over the supply of finished goods, and about how capability, once acquired, persists in a way that inventory does not.

Beijing may or may not have studied the Toshiba case. The policy structure suggests an understanding of the same principle.

One day before the October controls were to take effect, China suspended them until November 10, 2026. The July restrictions stayed in force. Fourth-generation LFP remains restricted.

I am not going to spend much time on the 2010 rare earth episode, because it has been referenced in every piece about Chinese export controls for the past three years and at this point it functions more as a rhetorical tic than as an analytical tool. The short version: China halted rare earth exports to Japan after a maritime dispute, never confirmed it, exports resumed, Japan spent a decade diversifying, the market permanently repriced supply risk. The battery situation has more regulatory infrastructure behind it and an additional feature that rare earths lacked: the licensing system generates transaction-level data on who is building battery capability, where, with which Chinese inputs. This is an intelligence function that operates independently of whether the controls are actively restricting supply.

Benchmark Mineral Intelligence tracked what happened after China restricted graphite exports in October 2023. South Korea got faster license approvals. India and the United States waited two to three months. Official policy stated equal treatment.

I want to dwell on this because it is the single most consequential operational detail in this entire story for procurement teams, and I have not seen a good explanation anywhere of how to actually manage around it.

A tariff can be modeled. A 25 percent tariff goes into the cost model and flows through to project economics. A licensing delay that might be two weeks or might be three months, depending on the buyer's country, depending on diplomatic conditions at the time of application, depending on criteria that MOFCOM does not publish, cannot be modeled in the same way. It can be hedged through inventory buffers, which tie up working capital, and through contractual protections, which Chinese suppliers resist because the protections shift risk to them. Most mid-size BESS developers and consumer electronics companies have not built either buffer into their operating models because, until 2023 with graphite and 2025 with cathode materials, there was no reason to.

Corporate procurement frameworks have line items for commodity price risk, logistics risk, supplier financial risk, quality risk. They do not have a line item for bilateral diplomatic risk affecting input material lead times. Building one requires organizational capability that most procurement functions do not have: the ability to assess political risk at a country level and translate it into supply chain planning parameters. Large automakers with government affairs teams can approximate this. A 200-person BESS developer in Texas cannot.

CATL's European plants, Erfurt and the upcoming Debrecen facility, produce cells and modules from cathode material shipped from China. Cathode synthesis stays in China. The Indonesia project, $6 billion, covers nickel extraction through cell production to recycling. 6.9 GWh initial capacity, scalable to 15 or 40 GWh. Full chain.

Europe gets assembly. Indonesia gets mine-to-cell integration. Indonesia has nickel, a cooperative government, no FEOC restrictions. Europe has regulatory complexity and governments that oscillate between courting and investigating Chinese battery investment depending on who won the last election.

There is a question I have been asking people in the European battery industry for months, and the answers I get are inconsistent enough to be worrying: does CATL Hungary carry its own cathode material inventory, or does it operate on just-in-time delivery from Chinese parent facilities? If it carries inventory, reimposition of controls gives buyers weeks or maybe a month of runway. If it runs JIT, the runway is days. Nobody I have spoken to outside of CATL itself seems to know the answer with certainty, and CATL does not disclose it. This is a specific, identifiable, answerable question about supply chain resilience for every European automaker sourcing cells from CATL's European operations, and the fact that the answer is not readily available tells you something about how seriously European buyers have been stress-testing their Chinese battery supply chains. Or how seriously they have not been.

I am going to keep this section shorter than it could be because the FEOC rules and IRA tax credit mechanics have been extensively covered elsewhere and I do not have much to add beyond what the trade press has already written. The summary: FEOC denies tax credits for vehicles with Chinese-controlled battery components. The restriction expanded to critical minerals in 2025. CATL is limited to arm's-length licensing in the U.S., as in the Ford BlueOval deal. Battery storage costs rose 56% to 69% through 2025 from tariff impacts. FEOC compliance phasing in through 2026 adds more.

The structural issue that deserves attention is the interaction between the two regulatory regimes. The U.S. system penalizes buyers for using Chinese components. The Chinese system restricts the technology transfers that would let buyers build alternatives. Non-FEOC cell makers available to U.S. buyers source anode materials and cathode precursors from China. Chinese export controls cover those upstream inputs wherever the downstream cell is assembled. Both systems make sense on their own terms. Where they intersect inside a buyer's supply chain, they create a problem that neither government designed and that has no obvious resolution forum because the two regulatory regimes do not interface with each other.

MOFCOM can require overseas entities to comply with Chinese export rules if those entities handle items produced abroad using Chinese-origin materials or technology. Exercised for rare earths already. Decision No. 61 of 2025 requires foreign entities to apply for licenses to resell certain Chinese-origin rare earth products even when the transaction occurs entirely outside China.

For batteries, the mechanism exists, the precedent is established, and I do not think anyone can confidently predict whether or when Beijing would use it. The practical consequence is that procurement departments are now doing supply chain origin audits they never did before. Anode material provenance was not a compliance variable until 2025.

John Goodenough's group at UT Austin developed LFP chemistry in the late 1990s. Published it. Japanese and Korean cell makers dismissed it for over a decade as a low-energy-density chemistry beneath their attention. Chinese manufacturers picked it up because it was cheap and safe, and then BYD and CATL, competing with each other through the 2020s, turned it into something the original inventors and the dismissive competitors never achieved. CATL's second-generation Shenxing adds 520 km of range in five minutes of charging. 800 km total range. 1.3 MW peak charging. Based on publicly available benchmarking, no non-Chinese cell maker has shipped LFP cells matching those numbers as of early 2026.

The export controls gate the transfer of this process knowledge and the equipment that implements it. Finished cell exports are unrestricted. CATL can ship from Ningde without limit. What requires a license is transferring the capability to replicate what happens at Ningde.

McKinsey estimated in early 2026 that Europe and North America combined hold five to ten percent of global production capacity across anodes, cathodes, electrolytes, and separators. I want to linger on that number because I think it has not sunk in for most people reading about battery supply chain diversification. Five to ten percent. After years of policy attention, billions in subsidies, hundreds of announcements about new factories. Five to ten percent.

The IEA put China's share of battery manufacturing above 80%. Over 90% of grid storage runs on LFP from China.

Artificial graphite is where diversification looks most hopeless. Over 90% of battery-grade supply from China. Graphitization furnaces run above 2800°C, consume enormous electricity, face environmental permitting barriers in Western jurisdictions. The North American Active Anode Material Producers lobbied Congress for a 920 percent tariff. That number has stuck with me since I first encountered it because of what it implies about the underlying cost structure. Even with generous assumptions about margin padding in the lobbying ask, it says the North American cost base is somewhere around ten times the Chinese cost base. The 25 percent tariff under discussion for January 2026 is a policy gesture. The gap it is applied against is an order of magnitude.

Sodium-ion eliminates lithium, nickel, and cobalt. The IEA noted in February 2026 that essentially all sodium-ion capacity, installed and announced, is in China. A chemistry change does nothing about the geography.

Nano One Materials in Canada has a cathode process called One-Pot that eliminates sulfate waste from conventional LFP synthesis. I flag this because sulfate waste disposal is specifically one of the reasons LFP production is expensive and hard to permit in Western jurisdictions, so a process that eliminates it addresses a specific barrier rather than the general cost problem. Nano One is at pilot scale. Chinese LFP production is at hundreds of thousands of tonnes. The distance between pilot and parity is years.

Solid-state batteries could bypass the graphite chokepoint through lithium metal anodes. The manufacturing challenges with solid electrolyte interfaces have proven stubborn. People working directly on solid-state development tend to put commercial-scale production after 2030 when they are speaking to peers rather than to investors.

China has nearly fifty graduate programs in battery chemistry and metallurgy. The U.S. has far fewer. This constrains every plan to build alternative manufacturing capacity because gigafactory ramp-up requires process engineers with production experience. The people who carry this knowledge overwhelmingly trained and gained factory experience in China. The export controls create uncertainty around what Chinese-trained engineers abroad may do, and the uncertainty is sufficient to slow cross-border hiring and push some candidates to decline offers. I do not have data on how large this effect is. Nobody does. It is the kind of thing that shows up as positions staying open longer and as candidates withdrawing late in the process, and that never gets aggregated into a statistic.

Global overcapacity hit around 900 GWh in 2025, per McKinsey. Pack prices fell to $108/kWh. The temptation for procurement teams is to conclude the export controls are a theoretical concern in a flooded market. The controls do not restrict cell exports below 300 Wh/kg. Chinese manufacturers with excess LFP inventory sell freely. What the controls restrict is the transfer of manufacturing technology and equipment. The market can be glutted with finished cells while the capability to produce those cells is being gated.

The $108/kWh price came from specific conditions: overcapacity from the 2021-2023 investment boom, aggressive competition, VAT export rebates subsidizing outbound shipments. China is phasing out those rebates. Section 301 tariffs are escalating. Several of these conditions are reversing at the same time. I do not think pack prices will return to $108/kWh on a sustained basis. McKinsey's own projections, along with Benchmark's, suggest stabilization at higher levels, though the specific forecasts vary and depend heavily on assumptions about Chinese domestic demand growth that nobody can make with confidence.

A procurement team accustomed to buyer's market leverage may read this as normal commercial hardening. Some of it is.

Force majeure clauses in battery supply contracts typically use generic language. Export control reimposition is foreseeable enough that generic language may not cover it. Name it as a trigger event.

Alternative supply commitments should specify non-Chinese-origin materials with audit rights going upstream. Chinese graphite processed through a Korean intermediary does not change its origin for purposes of Chinese extraterritorial claims.

DGAP recommended in December 2025 that European governments plan for China to extend restrictions to next-generation chemistries. The October controls, suspended, already contained a chemistry-agnostic 300 Wh/kg threshold. The template for broader controls exists. It has been drafted, published, briefly activated, and shelved. Procurement teams pricing supply agreements during this window are making a bet on whether the shelving is permanent. The trajectory across rare earths, critical minerals, and batteries over the past five years suggests it is not, and I think any procurement team that builds a three-year supply plan on the assumption that it is should be prepared to explain that assumption to their board.