America’s sanctions are now China’s Huawei sales pitch

Just a few months ago, Huawei’s chairman observed that Washington’s export controls have done little to hold China back. Instead, they have pushed the company and the Chinese state to invest far more heavily in domestic champions, accelerating an already rapid climb up the technology ladder and clearing the way for global expansion. The latest proof is in Egypt and Huawei’s offer to supply Cairo with roughly 2,000 artificial-intelligence chips. This deal is a signal that Huawei now feels confident enough to pitch its wares abroad, on the strength of a domestic manufacturing base it expects to keep scaling. Although this one deal won’t turn the country into an artificial intelligence powerhouse, it does expose a weakness in Washington’s export-control strategy: restrictions that slow Chinese companies at home can help them find customers abroad.Huawei has presented Cairo with a 12-month plan to develop AI capacity for military, surveillance, and other government applications. The proposal includes 1,408 top-end Ascend 950-series accelerators for training AI models and another 600 Ascend chips for running them. It also pairs Huawei with iFlytek, a Chinese surveillance company blacklisted by the United States. The computing power involved remains limited. These chips would provide performance comparable to only a few hundred of Nvidia’s best commercial accelerators. American companies operate far larger clusters. Egypt will gain useful AI capacity, though it won’t join the industry’s leading powers through this project. Washington’s restrictions have imposed real costs on Chinese companies. They deny access to cutting-edge processors and semiconductor manufacturing equipment, reducing efficiency and limiting Huawei’s ability to mass-produce accelerators. Huawei’s chips still trail Nvidia’s in speed, software compatibility, and energy efficiency. In a fast-moving industry, that advantage gives American companies valuable time. But export controls also change the long-term calculations of companies and governments. Chinese businesses that once preferred American semiconductors now have powerful reasons to develop domestic alternatives. Beijing has equally powerful reasons to finance them. Foreign buyers respond to the same pressures. Licensing delays and uncertain access prompt governments and companies to reconsider their purchases, engineering decisions, and software investments. Those choices create demand for alternative suppliers and direct money and technical talent toward improving their products. The first replacement often needs only to meet a customer’s immediate requirements. Customers, capital, and experience then improve its performance. Chinese AI company Z.ai offers evidence of that process. Its well-optimized GLM-5.3 model approaches leading Western models on some benchmarks while running efficiently on Chinese-made infrastructure. Egypt represents the next stage of Huawei’s strategy. The company previously developed Ascend processors primarily as domestic substitutes for Nvidia chips. It now seeks foreign customers. Its Egyptian offer combines processors, data centers, software, technical support, surveillance tools, and continued reliance on Chinese technology. Smaller markets offer Chinese suppliers an easier point of entry. U.S. semiconductor companies have traditionally treated Egypt as a lower-tier commercial opportunity, and U.S. rules have required licenses for advanced semiconductor equipment sales there since 2023. Huawei arrived with a state-backed package. Washington officials are now reportedly scrambling to assemble a counteroffer. That sequence should concern U.S. policymakers. Lengthy licensing reviews constrain American companies seeking to sell a U.S.-centered technology stack. Huawei can offer hardware, financing, software, and support as one package. Washington must then spend political and financial resources to recover customers that its own policies pushed toward a Chinese supplier. The Egyptian bid alone provides little evidence about the export-control system as a whole. Cairo may reject Huawei’s proposal or choose an American alternative, and Huawei still faces serious production constraints. The project’s importance nonetheless extends beyond the computing power inside one data center. An initial installation creates infrastructure, trains workers, establishes software standards, and produces maintenance agreements and political relationships. Each completed project makes later sales easier. Huawei followed this strategy in telecommunications, targeting markets that Western companies had overlooked and becoming a leading vendor across much of the developing world. The AI contest now encompasses the full technology stack that governments and developers adopt. Computing capacity, data center hardware, software, technical expertise, financing, and long-term support all shape purchasing decisions. Each new customer gives a supplier more revenue, operational experience, and developer feedback. Stricter U.S. regulations may therefore help Chinese companies improve their products and reach more foreign buyers. OPINION — REAGAN’S STAR WARS: HOW AMERICA STRIKES BACK AGAINST CHINA’S CRITICAL MINERALS EMPIREThe U.S. needs a faster, more predictable licensing process for American technology and services. Washington must also compete with the financing and technical support Beijing provides in developing markets. American companies have superior products, but licensing delays prevent them from serving some willing customers. Export restrictions can purchase time. That time has value only if the U.S. uses it to expand industrial capacity and secure foreign buyers. If Washington keeps American suppliers waiting while Beijing signs their customers, Huawei’s next proposal will come with larger orders and a longer customer list. Kristian Stout is director of innovation policy with the International Center for Law & Economics.