Are Chinese Cars Coming to America? What It Means for Ford and GM

Chinese cars are not, as of this writing, for sale in American showrooms. A 100%+ tariff on Chinese-made EVs, layered with additional Section 301 and Section 232 duties, has made that mathematically impossible — a BYD priced competitively in China roughly doubles in cost by the time it clears US customs. But “Chinese cars aren’t here yet” is a very different statement from “Chinese cars aren’t a threat,” and 2026 has been the year that gap became impossible for Detroit to ignore. And the loudest complication is that Washington itself can’t agree on what comes next.

 

 

The tariff wall makes the US case unusual, not the underlying demand. Look at what’s happened everywhere Chinese EVs have actually been allowed to compete on price. In Mexico, Chinese brands held roughly 12.5% of new-car sales in 2024; by early 2026 that had climbed to around 16%. In Brazil, BYD alone sold more than 37,000 vehicles in the first quarter of 2026, a 74% jump year-over-year, making it the country’s fifth-largest auto brand overall and helping push Chinese automakers past 80% of all EV sales nationwide. And in Australia — a market with no meaningful tariff barrier and no domestic auto industry left to protect — Chinese brands have gone from under 10% of new car sales four years ago to roughly 26% today, with BYD closing to within 243 vehicles of dethroning Toyota as Australia’s best-selling brand in June 2026 alone.

Why This Is Suddenly a Live Conversation

Three things converged this year. First, President Trump publicly floated welcoming Chinese automakers to build factories on American soil, provided they hire American workers — the same “transplant” model Toyota and Honda used decades ago. Second, Detroit’s own lobbying group, the American Automotive Policy Council, pushed back hard, arguing Chinese state subsidies and currency practices make that comparison unfair. Third, Ford’s own leadership has been unusually candid about the competition. CEO Jim Farley spent months driving a Xiaomi SU7 to understand it firsthand, and has described it as a high-quality car with a strong digital experience and very good performance. That candor isn’t new: as far back as 2024, Marin Gjaja, the COO of Ford’s EV unit, said Chinese EVs were “ahead of us in this technology,” and Bloomberg summarized Ford’s view as seeing low-cost Chinese EVs as a major strategic threat. Bill Ford has since gone further, warning publicly, as reported by the Wall Street Journal, that America can’t expect to keep Chinese EVs out forever and that Ford has to be able to beat them at their own game.

 Then Washington contradicted itself in public, in the span of a few days. On September 8, Transportation Secretary Sean Duffy sent Ford CEO Jim Farley a letter calling the company’s ties to Chinese firms “not a sustainable strategy for the United States” and urging it to cut them. The letter named Ford’s licensing deal with CATL for LFP batteries at its Marshall, Michigan plant, its Geely joint venture in Spain, reported hybrid-component talks with BYD, and its plan to keep the current Lincoln Nautilus in production in China until that generation’s run ends by the end of the decade, with the next-generation Nautilus moving to the US in 2030. Three days later, President Trump told Fox News he would be fine with Chinese automakers opening US factories as long as they employ American workers, comparing it to Japanese automakers building cars here. Ford’s reply made the whiplash explicit: it called the letter a “wrongheaded attempt to capture headlines,” and pointed out that the White House had praised the same Michigan battery plant days earlier and that Commerce Secretary Howard Lutnick had praised Ford’s decision to bring the next-generation Nautilus’s production to the US in 2030 the month before. Reuters characterized the week as Republican criticism and praise of Ford pulling in opposite directions, with the House Select Committee on China piling on as well.

This is a genuine split, not a messaging slip. Detroit News reported in February that many in the president’s own cabinet oppose Chinese automakers building in the US on economic and national security grounds, even after his January remarks that he’d welcome them. Observers have noted it isn’t obvious why Ford was singled out, given that GM’s US-built Chevy Bolt reportedly contains about 51% Chinese parts, including its LFP battery. And Ford is caught in the middle of its own paradox: it has argued it needs Chinese technology and expertise to stay competitive globally, even as it warns that Chinese automakers threaten the industry and should be kept out of the US. A CSIS analyst told Reuters the whiplash complicates automakers’ planning, and the practical result is real strategic uncertainty. The tariff wall on imports looks firm, but whether Chinese automakers can build here, and how much Chinese technology is acceptable inside American-built cars, is genuinely unsettled.

That’s the real story here: the debate has shifted from whether Chinese automakers can build competitive cars — that question is settled — to how and when they reach American buyers, with Washington itself still undecided on the answer.

The Real Root Cause: Corporate Culture Shapes the Requirements

Here’s the angle that gets missed in most coverage of this story: several of the technical advantages above — the wiring, the casting, the development speed — aren’t really the cause. They’re symptoms. (The battery advantage is a somewhat different story, rooted mostly in scale and supply-chain integration.) The actual difference is upstream, in how each kind of company generates its engineering requirements in the first place. That plays out in four connected steps: how requirements accumulate and why engineers rationally protect them, how an organization chart becomes a vehicle architecture, what a company without that history can do differently (and why different isn’t always better), and what it takes for different kinds of engineers to work together. The same pattern then shows up well beyond cars.

Requirements Accumulate for Good Reasons

A legacy automaker’s requirements database is not derived fresh from physics for each new program. It’s inherited. A modern OEM specification can trace individual requirements back decades — think of a wall-thickness minimum written after a field failure, a wiring redundancy rule added after a warranty problem, or a fastener torque spec carried forward from a platform that no longer exists. Each of these was a reasonable response to a real problem at the time. But once written into a spec, a requirement rarely gets re-derived from first principles — it gets carried forward, generation after generation, across all applicable vehicle architectures, because removing it requires someone to prove a negative: that the failure it was written to prevent can’t happen anymore. This is the same dynamic covered in our QFD and requirements cascade process — every requirement has to trace to something, but tracing forward is much easier than pruning backward.

It’s important to be fair to the legacy side here, because there is no shortage of engineering requirements at any major automaker, and the vast majority of them exist for good reasons. They weren’t handed down by a committee looking for something to do. They were built up over decades of real-world customer feedback and field data — warranty claims, failure analyses, and the hard lessons of vehicles that didn’t hold up the way their designers expected. Each requirement is, in effect, a scar from something that actually went wrong for a real customer. The job of a component or release engineer is to know those requirements and best practices inside and out for whatever part they’re responsible for, and to make sure the design meets them.

Here’s where the incentive structure matters. For an individual engineer, there is very little benefit to deviating from a requirement or a best practice, and a great deal of risk in doing so. If you follow the process and the part performs as expected, nobody notices. If you deviate and it works, the upside is usually modest — a little cost or weight saved. But if you deviate and an issue emerges in the field, the first question anyone asks is whether all the best practices were followed, because those practices exist precisely to prevent design issues. Follow them and the problem gets treated as something the process couldn’t have anticipated. Skip them and the problem is yours. As covered in our look at quality versus durability, the two aren’t the same thing — many factors shape a vehicle’s quality; however, if all best practices are followed, engineers will usually start looking at other factors — such as supplier quality, plant processes, etc. — as the likely culprit.

Put those two facts together and you get the outcome that shapes so much of large American automakers’ culture: engineers who are risk-averse, not because they lack talent or imagination, but because the reward structure makes that the rational choice. That is the mechanism behind requirements piling up — not timidity, but sensible people responding to the incentives they’ve been given. The real cultural challenge is building an organization where questioning a requirement with good data is safe enough to actually happen, without giving up the hard-won knowledge the requirements protect.

Do American Automakers’ Requirements Lead to Over-Engineered Cars?

The Cadillac Ashtray: Engineered for the Extreme

The most famous illustration comes from Bob Lutz, the veteran auto executive GM brought back in 2001, in his 2011 book Car Guys vs. Bean Counters (Lutz, 2011). Comparing the interior of a Cadillac STS against a similarly sized and priced Acura, Lutz touched the ashtray and it launched out of its recess under heavy spring pressure. Pushing it back in was uncomfortable enough to strain his fingers. The Acura’s tray, by contrast, glided open slowly and silently at a light touch and closed with a gentle, effortless push. A senior interior trim engineer was watching, and when Lutz asked whether he’d seen the difference, the engineer said he had, and that it was a proud moment for him. Cadillac’s was the only ashtray in the industry that met GM’s internal requirement, he explained, and no competitor, Mercedes included, met that standard. The standard was that any movable opening in a GM interior had to be fully functional after a night at minus 40°F.

Lutz then walks through the logic. Picture a driver in North Dakota at 5:30 on a minus-40 morning. If the doors aren’t frozen shut and the engine starts on a healthy battery, he might want a cigarette while the defroster clears the windshield, and he’d be appalled if the tray didn’t glide. As Lutz sums it up, GM had once again “engineered for an extreme situation,” and in doing so alienated thousands of customers in normal climates every day.

It’s hard to fault the engineer’s logic. The standard presumably traces back to real cold-weather failures, and the design passed it. But look at what the incentives rewarded: satisfying the standard. Nobody in that process was measured on how the ashtray felt to the far larger number of customers who would never see anything close to that temperature. The part had been optimized for a worst case at the expense of everyday use. Lutz held it up as an example of the obsolete design standards he found at GM.

The Mach-E vs. Model Y Teardown

A far more modern example shows the same tension in a system that matters a great deal more than an ashtray. In 2021, Sandy Munro and his Munro & Associates team, including Cory Steuben and Ben Lindamood, compared the thermal-management systems of the Ford Mustang Mach-E and the Tesla Model Y on their Munro Live series. Munro’s first reaction to the Mach-E’s system was to call it a nightmare, and the numbers from the teardown show why:

Measure	Ford Mustang Mach-E	Tesla Model Y	Mach-E vs. Model Y
Length of hoses	18.42 m	6.35 m	about 190% more
Part count	35 pieces	10 pieces	250% more
Fluid	22.4 kg	9.2 kg	about 143% more (13.2 kg extra)

 The Mach-E’s system uses nearly three times the hose length, three and a half times the parts, and more than twice the coolant of the Model Y. Autoevolution’s summary of the teardown adds that Ford needed two valves and four pumps to push fluid through those longer lines, where Tesla used a single Octovalve and two pumps. The consequences follow directly from the engineering. The extra 13.2 kg of coolant is mass the Model Y simply doesn’t carry. More hoses, valves, pumps, and fittings mean more parts to buy, ship, and assemble, which is why that same summary described Ford’s approach as heavier and costlier. And every additional hose, clamp, and connector is another potential leak path, so more failure points to validate, warranty, and eventually repair.

 This is also the same car whose wiring harness Ford later acknowledged was longer than it needed to be. But that’s an interpretation, not something the video proves. Munro also called the Mach-E the best EV his team had torn down other than Tesla’s, so this is a story about a gap to close, not a bad design. And a compact, integrated system has tradeoffs of its own, including how hard it can be to service.

The Takeaway: Requirements vs. Optimization

 So do requirements lead to over-engineered cars? They can, but the failure mode is more specific than simply doing too much engineering. It’s engineering to a worst case, or to an inherited architecture, without weighing who it serves and what it costs everyone else in mass, dollars, and failure points. The ashtray story also leaves a question open: it doesn’t tell us whether the Acura’s smoother mechanism would have survived the same cold soak, and better engineering may be able to satisfy both the extreme condition and the everyday feel.

 Both examples point to the same question. Any requirement or architecture inherited from the past can quietly become a constraint on everyone else, so the question worth asking isn’t only whether the design meets it, but who it protects and what it costs. Meeting a requirement and optimizing the system are not always the same thing.

When a Bigger Part Isn’t a Design Flaw

Over-engineering gets blamed for cost and mass problems, but those two failure modes aren’t automatic. The ashtray and the Mach-E’s cooling system got penalized for adding parts — more hoses, more valves, more pumps, more mass stacked on top of each other. Simply making one component bigger or more robust, without multiplying the part count around it, is a different decision entirely. As a rule of thumb, an oversized component tends to be more durable: there’s more material to wear through and more margin before it fails. Quality vs. Durability in Cars already covers how automakers weigh durability against cost at the system level — the same tradeoff plays out part by part.

Take two comparably sized vehicles with different brake sizing. The one with the larger rotors and calipers will almost always out-brake the other — shorter stopping distances, lower brake temperatures under repeated hard stops, more resistance to fade. None of that requires extra part count or added complexity; it’s simply bigger hardware doing more work. But bigger brakes cost more to produce, and that cost comes straight out of the vehicle’s margin. So why wouldn’t every automaker just fit the biggest brakes it can afford?

Look closely at a Land Rover lineup and the brakes often look larger than the vehicle’s everyday use case would demand. The easy read is over-engineering. But Land Rover models also tend to post some of the highest top speeds in their competitive segment, and a vehicle capable of a higher top speed needs proportionally more braking capacity to stop safely from that speed. The brakes aren’t oversized in isolation; they’re sized to a different requirement — the vehicle’s performance envelope, or its studio design target — not a lapse in judgment.

That’s the real point: making a part bigger is a decision, not a defect, as long as it isn’t dragging extra complexity along with it. The most cost-efficient brake system isn’t the same system as the highest-performing one or the most durable one, and no automaker optimizes for all three at once. Every manufacturer sets its own priorities — cost, durability, performance — and those priorities are a business choice as much as an engineering one. The automakers that get it right are the ones whose priorities trace back to what their customers actually value, rather than defaulting to whatever the spec sheet handed down.

When the Org Chart Becomes the Architecture

A second force is what’s sometimes called Conway’s Law: an organization’s products end up mirroring the organization’s own communication structure. A legacy automaker with a body team, an electrical team, a powertrain team, and a dozen supplier interfaces produces a vehicle architecture with exactly that many seams — a wiring harness that has to physically route through every one of those organizational boundaries, whether or not the underlying electronics actually require it. That’s part of why a traditional vehicle’s wiring harness can run several kilometers of copper: as much because of the org chart as because of the electrical requirement. The problem a major corporation faces is the silo effect, where the company is so large that the right hand doesn’t know what the left hand is doing, so the teams don’t always work together in the best interest of the vehicle as a whole, but instead optimize for their specific component or system at the expense of the whole vehicle.

Different Isn’t Always Better

Now compare that to a company that didn’t inherit fifty-plus years of automotive requirement sediment. Many Chinese EV makers weren’t automakers at all a decade ago — they were battery, electronics, or software companies that decided to build a car. With far less legacy to defend, the natural (and much cheaper) question becomes: what does the actual physics require, right now, with today’s materials and today’s electronics — not what did a 1998-era wiring architecture require? That’s the literal definition of first-principles engineering: deriving a requirement from the underlying physical constraint (how much current has to move, how far, at what voltage drop tolerance) instead of from precedent. It’s the same reasoning behind the newest zonal wiring architectures, which cut harness length dramatically, and it’s a big part of why gigacasting works: fewer discrete parts means fewer discrete, organizationally owned specifications standing in the way of a simpler design.

But the legacy side isn’t standing still, and this is where the story gets more interesting. Ford’s Universal EV program is the clearest example of a legacy automaker challenging its own requirements. Ford set up a small skunkworks team in California, deliberately away from the checks, processes, and bureaucratic steps that had built up over decades, and asked it to rethink the vehicle starting from the function it needs to provide. Detroit News reporting describes a platform with about 20% fewer parts, 25% fewer fasteners thanks to large unicastings, and a zonal electrical architecture whose wiring harness is roughly 4,000 feet shorter and 22 pounds lighter than that of Ford’s first-generation EV. It’s the same lesson Farley said Ford learned in 2023: the Mustang Mach-E’s wiring harness turned out to be about 1.6 kilometers longer than it needed to be, adding weight and roughly $300 in extra battery cost (mass compounding at its finest). S&P Global Mobility described the effort as a legacy automaker challenging long-standing industry norms. Legacy automakers, in other words, are already aggressively questioning the requirements that no longer make sense.

Purging outdated requirements isn’t new for legacy automakers, either. Lutz describes GM doing exactly that in the 2000s: he says the engineering culture was full of inherited do’s and don’ts, hundreds if not thousands of them, and a special team led by an executive he calls Captain Queen met at 6:00 every morning and systematically eliminated about 90 percent of these requirements, which no longer made sense. Automakers periodically clear out old requirements. What’s different now is that Chinese competition is raising the urgency.

Meanwhile, the newcomers still have hard lessons ahead of them, and Xiaomi’s SU7 is a useful case study. Xiaomi is a consumer-electronics company that only recently started building cars, and its sedan launched with real momentum. By May 2025, though, new SU7 orders had fallen 55% in April from March, according to Reuters reporting summarized by EV.com. The drop came amid an investigation into a fatal accident involving the car, which put the safety of Xiaomi’s smart-driving features under public scrutiny and led Chinese regulators to tighten oversight of how such features are marketed. Around the same time, nearly 400 SU7 Ultra owners had asked for refunds over an optional 42,000 yuan carbon fiber hood that they said lacked the internal ducts it had been described as having, and CEO Lei Jun acknowledged it was the most difficult month since he founded the company. The Ultra’s slide continued: CarNewsChina reports that its monthly sales, which ran between roughly 2,000 and 3,000 units from March to August 2025, fell to just 45 units in December, amid further controversies that included false-advertising complaints, advance-payment requirements, and reports of doors failing to open after an accident.

It’s worth being fair about cause and effect. CarNewsChina says only that the controversies may have affected the Ultra’s sales, and the Ultra is a niche performance model. Xiaomi’s overall business kept growing, with 411,837 vehicles sold in 2025 (up roughly 201%) and the standard SU7 still selling 11,123 units in December. The lesson isn’t that Chinese EVs are unsafe or that the newcomers are failing. It’s that driver-assistance safety, honest claims about optional equipment, and what happens to occupants after a crash are exactly the areas where hard lessons produce requirements, and a company moving fast tends to meet those lessons in public — Tesla is notorious for doing this.

The tradeoff runs both ways. A newer entrant with no decades-long warranty history has fewer precedents to defend and can move faster, but it also hasn’t yet paid for the lessons those requirements encode, and it likely doesn’t have enough requirements yet. Dropping a requirement without understanding why it exists can mean learning the same lesson again, this time from your own customers. I see the industry landing in a middle ground. Chinese competition pushes legacy automakers to become more efficient, clearing out requirements that no longer make sense, as GM did under Lutz and as Ford’s Universal EV team is doing now. At the same time, Chinese automakers, and their regulators, write more stringent requirements as real-world data and hard lessons accumulate. The playing field evens out from both directions. Different isn’t always better. What wins is knowing which requirements to challenge and which ones to keep.

Physics Is the Same for Everyone

Both Chinese and American OEMs are bound by exactly the same laws of physics. I’ve personally met extremely talented engineers on both ends of a real spectrum — deep subject-matter experts who understand one narrow piece of a vehicle better than almost anyone else in the building, and broad, big-picture thinkers who know less about any single component but are exceptionally good at seeing how a change to one part will cascade across a dozen other systems they don’t personally own. Neither type is more valuable than the other, and no vehicle program succeeds without both. The real job of corporate culture is getting those two very different groups of engineers to actually collaborate instead of working past each other, to produce a vehicle that’s high quality, visually appealing, packed with the latest technology, and still priced at something an actual customer can afford. That’s a genuinely difficult four-way tradeoff, and it’s worth being honest that no company, in any country, fully solves it. What separates the winners right now isn’t raw engineering talent on either side — it’s whether an organization’s culture gets out of the way of that collaboration, or adds friction to it.

This Isn’t an Isolated Automotive Story

The same broad pattern — an incumbent carrying accumulated precedent and cost structure, versus a challenger that is free, or forced, to rebuild from first principles — shows up across nearly every technology sector where China has rapidly closed a cost gap. The drivers differ from sector to sector (freedom from legacy, constraint that forces efficiency, state support, sheer scale), but the direction is the same:

• Artificial intelligence — Chinese lab DeepSeek’s R1 model was reported to have cost roughly $6 million for its final training run (a figure that excludes earlier research and hardware costs), yet matched the performance of Western frontier models that cost far more to develop. Independent benchmarking since has found several Chinese models completing a standardized task for between two and 33 cents, versus roughly $2.75 for one leading Western frontier model. Here the driver is closer to constraint than to freedom from legacy: facing chip export restrictions, Chinese labs were pushed to squeeze more out of every unit of compute through model architecture and inference efficiency.
• Robotics — five Chinese manufacturers (AgiBot, Unitree, Galbot, UBTECH, and Leju Robotics) accounted for roughly 86% of global humanoid robot shipments in the first half of 2026, at price points reported to undercut Western competitors dramatically. Unitree in particular has pushed to bring more of its component supply chain in-house rather than inheriting the layered supplier relationships common to Western robotics firms.
• Space — the United States and China are engaged in a modern space race focused on lunar exploration, satellite networks, and orbital dominance, with both nations planning to land astronauts on the Moon before 2030.
• Aviation — COMAC’s C919 airliner has been priced at roughly $90–110 million by most estimates, compared with around $111 million for an Airbus A320neo and $121 million for a Boeing 737 MAX.

Not every one of these is a clean first-principles story, and the mechanism differs by sector. But the recurring advantage isn’t one country’s cheaper labor or a single clever trick: it’s a structural freedom, sometimes chosen and sometimes forced, to ask what a solution actually requires rather than what it has always required. None of this means the legacy approach is simply wrong. The requirements encode genuinely hard-won knowledge, and the engineers who defend them are responding rationally to the incentives they’ve been given. But it does mean the real competitive gap isn’t primarily a battery chemistry or a casting technique. It’s that some organizations face far fewer obstacles to asking “why does this requirement exist at all?” — and the ones carrying decades of precedent have to find a way to ask it safely.

What This Means for American Automakers

The pressure is already visible in the numbers, tariffs or not. GM and Ford together have taken more than $27 billion in EV-related write-downs, canceling models and slowing production as demand cooled faster than planned. Chinese brands held roughly 16% of Mexico’s new-vehicle sales in early 2026 — 22.8% counting every vehicle built in China regardless of brand — in a USMCA neighbor with much lower tariff barriers than the US. That puts real, geographically close competitive pressure on Detroit even without a single Chinese car crossing into the US.

 

And it isn’t limited to Mexico. More and more of the markets Ford and GM actually sell into are opening their doors to Chinese competition. Europe absorbed roughly 811,000 Chinese-brand vehicles in 2025 alone, pushing China’s share of that market to about 6.1% and still climbing, with BYD’s European market share now exceeding Tesla’s. Canada, historically aligned with US trade policy on this exact issue, has cracked its own door open: as of January 2026, it allows up to 49,000 Chinese EVs in per year at a preferential 6.1% duty — a fraction of the wall the same vehicles face crossing into the US. Add the ground already lost in Mexico, Brazil, and Australia, and the picture is clear: even if the US tariff wall never moves an inch, Ford and GM still end up competing directly against Chinese automakers in Europe, Canada, Latin America, and Australia — markets that matter to their global bottom line regardless of what happens at the US border.

There’s also a possible back door: Geely (which owns Volvo and Polestar) is seen by industry analysts as the Chinese automaker best positioned to eventually build in North America, potentially through a USMCA-qualifying plant that would sidestep the tariff wall entirely — the same strategy BYD has evaluated in Mexico. If that happens, “Chinese cars in America” stops being a hypothetical and becomes a supply chain and labor question overnight, which is exactly why the United Auto Workers’ interests don’t fully align with Detroit’s own lobbying position on this: a non-union transplant plant helps no one on the labor side, tariff wall or not. Geely is also Ford’s new partner in Spain, as covered next, which is a big part of why the line between competitor and partner is so blurry here.

We’ve Seen This Movie Before: The Japan Parallel, and Ford’s Answer in Valencia

None of this is actually unprecedented. American automakers lived through a nearly identical playbook once already, with Japan — and the eventual resolution wasn’t containment. It was partnership.

Through the 1970s, Toyota, Honda, and Datsun (later Nissan) built their US reputation the same way Chinese automakers are building one today: undercutting Detroit on price while matching or beating it on reliability, right as the 1973 and 1979 oil shocks made Detroit’s large, thirsty sedans look like a bad bet. Detroit’s initial response was dismissal, then lobbying. In 1981, the US and Japan negotiated a Voluntary Export Restraint capping the number of Japanese cars that could be imported each year — the tariff-style protection playbook of its day. It worked about as well as protection usually does: it slowed the bleeding without stopping it, and it gave Japanese automakers a strong incentive to stop exporting and start building here instead. Honda opened a plant in Marysville, Ohio in 1982. Nissan followed in Smyrna, Tennessee in 1983. Toyota partnered with GM on the NUMMI joint venture in Fremont, California in 1984, then opened its own standalone Georgetown, Kentucky plant in 1988. Within a decade, “Japanese cars” had quietly become, in large part, American-built cars engineered to Japanese standards — and as our own look at why Japanese vehicles last longer covers, that quality gap never fully closed. Detroit didn’t defeat that wave. It absorbed it, later than would have been ideal, and largely on the newcomer’s terms.

Ford is already running a version of that same play with China, in real time. On July 23, 2026, Ford and Geely announced a joint venture to build vehicles together at Ford’s Valencia, Spain plant — a nearly 50-year-old facility with roughly 500,000 units of annual capacity that produced under 100,000 vehicles in 2025. The arrangement, 66% Ford and 34% Geely, will put an all-new Ford multi-energy crossover and a new Bronco family member on the same production line as two electric Geely-branded SUVs, with output starting in 2028 and built in part on Geely’s own Global Intelligent New Energy Architecture platform — the same underlying cost and engineering advantage discussed throughout this piece. In plain terms: Ford is buying access to Geely’s platform economics for an underused European plant, and Geely is buying a tariff-clean route into the EU market alongside the Volvo, Polestar, and Lotus brands it already owns there. It’s a notable twist that the partner is Geely — the same Chinese automaker widely seen as best positioned to build in North America, and one whose tie-up with Ford Transportation Secretary Duffy specifically named in his September letter. Neither side is calling it what the 1980s version would have been called — protection eventually turning into partnership — but the shape of the deal is unmistakably familiar.

The clearest historical echo of this specific move is NUMMI. When Toyota and GM opened that Fremont, California joint venture in 1984, the real product wasn’t just the badge-shared cars coming off the line — it was the Toyota Production System itself. GM rotated waves of its own managers through NUMMI specifically to learn lean manufacturing firsthand, and what they brought back reshaped how the rest of the company built cars for decades afterward. Valencia isn’t a repeat of that story in every detail: Ford already knows how to build cars, and this deal is an ownership stake rather than a training exercise. But the underlying logic rhymes. An American automaker is deliberately sharing a factory floor with a foreign partner specifically to absorb a manufacturing approach it doesn’t fully have in-house, at a plant that may end up defining an era well past the deal that created it.

The Bottom Line

Nothing here means BYD is arriving at a dealership near you next year. The tariff wall on imports is real, bipartisan, and not going anywhere soon. What’s unsettled is whether Chinese automakers get to build here, and on that Washington is openly split: the same administration is criticizing Ford for its Chinese partnerships while the president says he’d welcome Chinese-owned plants. But the underlying competitive gap — in battery cost, manufacturing technique, and development speed — is the thing Detroit actually has to solve, tariffs or no tariffs. Ford’s own leadership is already saying so publicly. And because Ford and GM already compete against Chinese automakers in Europe, Canada, Latin America, and Australia whatever happens at the US border, this isn’t a fight they can opt out of just because they’re currently winning it at home. That, more than any single headline about a factory announcement, is the part of this story worth watching.

Frequently Asked Questions

Are Chinese cars sold in the US right now?

No. As of 2026, tariffs exceeding 100% on Chinese-made EVs have made Chinese-branded passenger vehicles commercially unviable in the US market. No major Chinese automaker currently sells passenger cars here.

What happens when Chinese EVs enter a market without tariff barriers?

Market share grows very fast. In Mexico, Chinese-built vehicles reached 22.8% of sales through July 2026. In Brazil, Chinese automakers now account for over 80% of EV sales, with BYD alone becoming a top-five overall brand. In Australia, Chinese brands went from under 10% of the market four years ago to roughly 26% today, with BYD closing in on Toyota as the country’s best-selling brand.

Has America seen a similar situation before?

Yes. In the 1970s and 80s, Japanese automakers like Toyota, Honda, and Nissan undercut Detroit on price and quality, prompting a 1981 Voluntary Export Restraint. Rather than blocking Japanese automakers permanently, that restraint pushed them to build US “transplant” factories instead, including Honda’s 1982 Marysville, Ohio plant and Toyota’s 1988 Georgetown, Kentucky plant — a pattern of protection turning into partnership that some see echoed in deals like Ford and Geely’s 2026 Valencia, Spain joint venture. Toyota’s NUMMI joint venture with GM in Fremont, California is an even closer parallel: GM used it specifically to learn Toyota’s lean manufacturing system, and the same Fremont plant is now Tesla’s factory.

What is the Ford-Geely Valencia joint venture?

According to Ford’s official announcement, Ford and Geely Auto agreed on July 23, 2026 to form a joint venture at Ford’s Valencia, Spain plant, with Ford owning 66% and Geely Auto 34%. Pending regulatory approval, the joint venture will begin operations in the first half of 2027; the Ford Kuga will keep being built at Valencia in the meantime. The first new vehicles are due in 2028: an all-new Ford multi-energy family crossover jointly developed with Geely, a new member of the Bronco family, and two electric Geely-branded SUVs. Ford says the deal is part of a wider push to bring five new multi-energy vehicles to European showrooms by 2029, and traces the partnership back to 2010, when Ford sold Volvo Cars to Geely. Valencia has been in operation since 1976, when it built the original Ford Fiesta, and Ford describes its potential annual capacity as about 500,000 vehicles.

Could Chinese automakers build cars in the US to avoid tariffs?

It’s been discussed. President Trump has said he’d welcome Chinese automakers building US factories with American workers, though Detroit’s own lobbying group has pushed back. A more likely near-term path is a Mexico-based, USMCA-qualifying plant — an approach BYD has evaluated.

Is the Trump administration for or against Chinese automakers building in the US?

It has sent conflicting signals. In a September 11 Fox News interview, President Trump said he would accept Chinese automakers opening US factories if they hire American workers. Three days earlier, on September 8, Transportation Secretary Sean Duffy sent Ford a letter urging it to cut ties with Chinese companies including CATL and Geely, citing national security concerns. Reporting indicates many in the cabinet also oppose Chinese automakers building here, and the Detroit Three’s lobbying group, the American Automotive Policy Council, opposes the idea as well.

Why are Chinese EVs so much cheaper to build?

A mix of factors: LFP battery chemistry and scale that make batteries cheaper than what most Western automakers use, gigacasting techniques that reduce part count and assembly labor, simplified wiring architectures, and development cycles roughly half the length of traditional Western timelines — on top of state subsidies. Underlying much of it is an organizational difference: many Chinese EV makers have far less legacy to defend and can derive specs more directly from first-principles physics rather than historical precedent.

Is this cost-advantage pattern unique to cars?

No. Similar cost gaps and competitive races show up in AI (DeepSeek and other Chinese labs completing tasks at a fraction of Western frontier-model costs), robotics (Chinese firms accounted for about 86% of global humanoid robot shipments in the first half of 2026), space (the US and China both targeting a crewed Moon landing before 2030), and aviation (COMAC’s C919 is priced below Airbus and Boeing, though it is heavily state-subsidized). The drivers differ — freedom from legacy requirements, constraints that force efficiency, state support, and scale — but the direction is the same.

Are Ford and GM competing with Chinese automakers even without a US market entry?

Yes. Chinese automakers are rapidly gaining share in markets Ford and GM depend on globally, including Europe (roughly 6.1% share and rising in 2025), Canada (which now permits a quota of Chinese EVs at a preferential tariff rate), Mexico, Brazil, and Australia. Detroit doesn’t need a US market entry to feel this competitive pressure — it’s already facing it abroad.

How are Ford and GM responding?

Both have taken major EV write-downs (over $27 billion combined) while slowing their EV rollouts. Ford has also gone further than write-downs, agreeing to a 2026 joint venture with Geely to build vehicles together at its Valencia, Spain plant, and CEO Jim Farley has been unusually public about treating Chinese EVs as a serious engineering benchmark rather than dismissing them.

References

1. Lutz, Bob. Car Guys vs. Bean Counters: The Battle for the Soul of American Business. Portfolio/Penguin, 2011. Publisher page.
2. Munro Live (Munro & Associates). “Mach-E Thermal System Part 2: A Detailed Comparison to the Model Y.” YouTube, July 2021. https://www.youtube.com/watch?v=m1kHsd3Ocxc
3. EV.com Staff. “China’s Tesla Model 3 Rival Xiaomi SU7 Sees Sales Drop Amid Consumer Blacklash.” EV.com, May 15, 2025. https://ev.com/news/china-tesla-model-3-rival-xiaomi-su7-sales-drop-consumer-blacklash
4. Liu Miao. “Xiaomi SU7 Ultra sales plummet: from over 3,000 to just 45 units monthly.” CarNewsChina, January 31, 2026. https://carnewschina.com/2026/01/31/xiaomi-su7-ultra-sales-plummet-from-over-3000-to-just-45-units-monthly/
5. “How Ford’s ‘skunkworks’ team is designing a more efficient EV.” The Detroit News, February 17, 2026. https://eu.detroitnews.com/story/business/autos/ford/2026/02/17/how-fords-skunkworks-team-is-designing-a-more-efficient-ev/88715496007/
6. “Ford’s all-in bet on automotive innovations.” S&P Global Mobility, August 28, 2025. https://www.mobilityglobal.com/en-us/automotive-insights/blog/fords-all-in-bet-on-automotive-innovations
7. Ford Motor Company. “Ford and Geely Auto Join Forces in Europe to Produce Next-Generation Multi-Energy Vehicles in Spain.” Ford From the Road, July 23, 2026. https://www.fromtheroad.ford.com/us/en/articles/2026/ford-and-geely-auto-join-forces-in-europe
8. Christensen, Jen, and Kristin Fisher. “What we know about China’s moon-landing plans and why the US wants to beat them there.” CNN, August 7, 2026. https://www.cnn.com/2026/08/07/science/china-us-moon-space-race