Ford took a beating in 2023 for licensing battery technology from China’s CATL. Now The Information is making the opposite case: Ford was right, and the rest of the auto industry should follow its lead. It’s a contrarian take on one of the most politically charged supply chain decisions of the decade, and the argument lands harder today than it would have three years ago.
Here’s the setup, for anyone who missed the original fight. Ford is building a battery plant in Marshall, Michigan that produces lithium iron phosphate (LFP) cells using technology licensed from CATL, the world’s largest battery maker. Ford owns the plant and employs the workers. CATL gets royalties and provides the know-how. Lawmakers called it a Trojan horse. Ford called it the fastest way to get cheap, durable batteries onto American soil.
What the argument actually is
The Information’s core claim, as the headline frames it, is straightforward: Chinese companies are years ahead on LFP chemistry and manufacturing, and pretending otherwise costs Western automakers time and money they don’t have.
A few facts make that hard to dispute:
- LFP batteries skip nickel and cobalt entirely. That means lower cost, fewer supply chain headaches, and cells that tolerate thousands of charge cycles without much degradation.
- CATL and BYD together control the majority of global LFP output. Western suppliers are still ramping.
- Ford’s cheap EV plans, including the sub-$30,000 midsize pickup it’s building on a new platform, depend on those cells. Without LFP at scale, the math on affordable EVs doesn’t close.
What stands out here is the licensing structure. Ford didn’t hand CATL a stake or let it operate the factory. It bought the recipe and kept the kitchen. That’s the model The Information is pointing to as the template.
Why this matters now
This piece isn’t just about cars. The same debate is playing out across the AI industry, and the parallels are uncomfortable for anyone who wants a clean “buy American” answer.
Chinese open-weight models from DeepSeek, Alibaba’s Qwen family, and Moonshot have become default choices for plenty of Western startups because they’re good and they’re cheap to run. Companies fine-tune them, host them on U.S. infrastructure, and ship products. That’s a licensing relationship in everything but name. The Ford question, whether it’s acceptable to build on Chinese technology when Chinese technology is ahead, is the same question a CTO faces when picking a base model.
The pushback is real on both fronts. Washington has spent the past two years tightening rules on foreign entities of concern in the EV tax credit program, and there’s ongoing pressure to restrict Chinese AI models in government and defense work. So the pragmatic path carries regulatory risk, and that risk moves with each election cycle.
Where this goes over the next few years
My read, projecting out to 2028 or so:
- Licensing becomes the norm, not the scandal. If Ford’s Marshall plant hits its cost targets, expect GM, Stellantis, and the Japanese automakers to quietly strike similar deals. Some already have partial versions. The political heat fades once the jobs show up.
- The line shifts from “where is it from” to “who controls it.” Regulators will increasingly ask whether the IP, the data, and the operations sit under domestic control. Ford’s structure passes that test. A joint venture with a Chinese operator wouldn’t. AI companies should expect the same framing applied to models and weights.
- Domestic alternatives catch up, but slowly. Western LFP suppliers will get there, and so will Western open-weight models. Licensing buys the years in between.
Practical takeaways
If you’re running a company that touches hardware or AI infrastructure:
- Separate the technology from the vendor. Ask what you’d own if the relationship ended tomorrow. Ford would still have a working plant and trained staff. Make sure you’d still have your weights, your data, and your deployment.
- Document provenance now. Whatever you build on, keep a clear record of where it came from and under what terms. That paperwork will matter when the rules change.
- Don’t confuse political discomfort with technical risk. They’re different problems with different mitigations.
The Information’s full piece has more detail on Ford’s economics and the reasoning behind the recommendation. Whatever you think of the politics, the question it raises isn’t going away. It’s just going to show up in more industries.