Selling AI Chips Won’t Keep China Hooked on U.S. Technology

In August, the Trump administration reversed course on Nvidia’s H20 AI chips, lifting previous restrictions under a 15 percent revenue-share condition, and is now considering allowing exports of a slightly downgraded version of Nvidia’s Blackwell — America’s most advanced AI chip. Proponents argue it will keep Chinese developers “addicted to the American technology stack” because Nvidia’s technology is hard to replace or replicate. The evidence suggests otherwise: in the long term, selling U.S. AI chips to China is unlikely to create lasting dependencies on the American tech ecosystem.

At the heart of the “addiction” theory lies the belief that using American chips creates vendor lock-in, compelling foreign AI developers to remain within U.S. tech platforms and to keep buying American chips. These arguments have some merit. Nvidia’s proprietary networking equipment and stack strongly incentivizes engineers to stay within its ecosystem, having refined its CUDA software platform over nearly two decades into a “strategic moat.” With its vast libraries of pre-written code and supporting tools, CUDA allows developers to leverage the parallel computing capabilities of Nvidia chips. When AI companies build on and develop expertise in Nvidia’s hardware and software ecosystems, it creates a degree of path dependence. This lock-in makes it more likely that each additional dollar of AI investment flows to and strengthens the American — rather than Chinese — AI ecosystem, reinforcing the U.S. lead. Proponents use this argument to justify selling AI chips to China: get the country’s tech sector “hooked” on Nvidia to capture greater market share and redirect Chinese investment toward U.S. AI innovation.

But making China dependent on U.S. technology is not that straightforward. AI chips are more akin to generators than utility companies. Generators are necessary to produce power, but once the generator is running, the manufacturer does not control what is powered with it. Generators can be swapped out for others, combined with local power sources, or integrated into hybrid energy systems. Unlike a utility company that maintains persistent control over electricity supply and pricing, chips are a one-off input — value-neutral hardware that runs whatever code developers choose. Developers may use CUDA today, but they can layer domestic software and tools on American hardware tomorrow. They can gradually integrate these systems with local infrastructure as their domestic ecosystem matures. Beijing has a well-funded national strategy to indigenize chip production, and access to U.S. chips will not meaningfully diminish these efforts. Rather than creating lasting dependence, exporting U.S. chips will simply expedite China’s AI progress as it scales its indigenous chip manufacturing capacity.

The United States’ own AI labs show that reducing reliance on Nvidia’s proprietary ecosystem is not only possible, but already happening. Anthropic originally relied on Nvidia graphics processing units (GPUs) to train its flagship model, Claude, but it has since shifted most of its computing needs to other ecosystems. Today, Anthropic optimizes for training on AWS’s Trainium hardware, while also using a mix of Google tensor processing units (TPUs) and Nvidia GPUs. Google DeepMind followed a similar trajectory. The company once depended on Nvidia GPUs for early breakthroughs, but it has since moved beyond CUDA and now trains its capable Gemini models on Google TPUs. Both shifts highlight a crucial lesson: Nvidia’s chips are less of an addiction than a gateway drug. There is clear proof that the Nvidia moat is surmountable. If U.S. AI labs with no political mandate to leave the Nvidia stack can accomplish this transition, then Chinese labs facing clear national incentives can likely do the same.

Read the full article on Just Security.