How ZTE changed Xi Jinping’s vision for China’s tech future

(Edited and refined by Candice Chan, with the assistance of AI translation.)

The 2018 US sanctions crisis that nearly crippled ZTE was a pivotal moment in the evolution of China’s economic and technological strategy. It laid bare a fundamental reality: being a global manufacturing powerhouse is not the same as possessing technological sovereignty. China might be capable of producing sophisticated telecommunications equipment on a vast scale, but a single decision by Washington was enough to bring one of its leading technology companies close to a standstill within weeks.

For Chinese President Xi Jinping, the crisis also came as a profound political and psychological shock. China’s dependence on American technology was so great that Xi himself had to appeal to Donald Trump for help, hoping the US would give ZTE a reprieve. The ZTE affair therefore became much more than a corporate dispute. It was a stark lesson in national power, strategic vulnerability, national dignity and the risks of technological dependence.

The strategic evolution that followed can be summed up as:

ZTE shock → technological humiliation → security consciousness → technological self-reliance → priority for advanced manufacturing → open technology ecosystems → sovereign artificial intelligence

The ZTE affair: China’s ‘technological Sputnik moment’

The ZTE crisis began with violations of US sanctions against Iran and North Korea, but what truly alarmed Beijing was Washington’s demonstrated ability to exploit technological choke points.

In April 2018, the US Department of Commerce banned American companies from supplying ZTE. The company immediately lost access to critical technologies, including Qualcomm processors, Google’s Android ecosystem, FPGA chips and electronic design automation (EDA) software. Before long, it was forced to suspend its main business operations.

China was confronted with a sobering reality: a technology giant employing tens of thousands of people could effectively be switched off at the touch of a button because a handful of critical technologies were controlled by the US.

It was China’s “technological Sputnik moment”. Technological autonomy, previously a long-term goal of industrial development, suddenly became a matter of national security.

The ZTE crisis exposed the difference between industrial scale and technological sovereignty. China had already become a global industrial power, yet critical areas such as semiconductors, operating systems, design software and intellectual property remained subject to US control. These were the vulnerabilities that China would later repeatedly describe as “choke-point technologies”.

Xi and the ZTE reckoning

The ZTE affair took on particular political significance because Xi Jinping intervened personally.

In May 2018, Xi appealed directly to Trump for help in preventing ZTE’s collapse. Trump subsequently instructed the Commerce Department to seek a solution.

But the terms imposed by the US underscored the imbalance of power between the two sides. ZTE ultimately incurred about US$2.29 billion in civil and criminal penalties, was forced to replace its entire board and senior management, and had to submit to stringent US compliance oversight. Most symbolically, it was required to accept US-approved monitors inside the company for ten years, with ZTE itself bearing the cost.

For Xi, whose political mission centres on the “great rejuvenation of the Chinese nation”, the message could hardly have been clearer: China might have achieved an extraordinary economic rise, but the fruits of that development could still be held hostage to an American technological veto.

The psychological and strategic impact of the ZTE affair should therefore not be underestimated. Washington might have viewed it as an exercise in export-control enforcement, but Beijing saw a much larger danger: economic interdependence could be weaponised.

From development first to ‘security and development’

China had been promoting indigenous innovation well before 2018. The ZTE affair did not create its drive for technological self-reliance. Rather, the equation was:

Existing technological self-reliance strategy + ZTE shock = acceleration, securitisation and political mobilisation.

Before ZTE, dependence on foreign technology could largely be treated as an economic question. If imported technology was more advanced and cheaper, it made sense for China to buy it. After ZTE, technological dependence was increasingly viewed as a national security issue.

Once technology was brought within the national security framework, the economic calculus changed as well. Domestic substitution acquired strategic value even if it was more expensive in the short term. Redundancy in supply chains was no longer simply a source of inefficiency, but a form of insurance. Supply chain diversification, technical standards, intellectual property, software platforms and technological ecosystems all became components of national power.

Xi later encapsulated this logic in the phrase: “Security is a prerequisite for development, and development serves as a safeguard of security.”

Technology and advanced manufacturing over macroeconomic stimulus

This strategic shift also helps explain an apparent contradiction in China’s economic policy today.

With the property sector undergoing a prolonged adjustment, consumption remaining weak, deflationary pressures persisting, and private-sector confidence subdued, there is a clear case for stronger macroeconomic stimulus. Yet Beijing continues to devote enormous political and financial resources to semiconductors, AI, robotics, advanced manufacturing, industrial upgrading and technological self-reliance.

From the perspective of conventional macroeconomics, this allocation of resources may seem unbalanced. But seen in light of the strategic lessons Beijing drew from the ZTE crisis, the logic is clear.

The policy objective has gradually shifted from maximising economic efficiency alone to ensuring that China has sufficient national resilience to withstand prolonged US-China competition and geopolitical pressure.

In other words, Beijing is concerned not only with whether consumption grows by a few percentage points today. It is also asking a more fundamental question: if US-China relations deteriorate further in the future, will China still have access to the critical chips, software, equipment and artificial intelligence technologies it needs?

Technological self-reliance does not mean shutting China off from the world

China’s answer is not to build a technological Great Wall and recreate every Western technology at home. Such an undertaking would not only be enormously expensive; it could also isolate China from global knowledge networks and ultimately weaken its capacity for innovation.

Instead, China has effectively adopted a strategy of “strategic openness”: maximising access to global knowledge while minimising its vulnerability to technological coercion by foreign powers.

This gives open-source technologies particular strategic value. Open ecosystems allow Chinese companies and engineers to inspect, modify, adapt and maintain technologies themselves, reducing their dependence on any single proprietary foreign supplier.

China can therefore pursue three objectives at once:

Participation in global technology networks + domestic substitution + strategic redundancy

This is neither the globalisation of the past nor complete decoupling. It is a form of “globalisation with an insurance policy”.

For instance, China has developed domestic alternatives such as Gitee while continuing to participate in GitHub. Beijing does not want Chinese researchers to withdraw from global technology networks. Rather, its goal is to remain part of the global system without becoming hostage to it.

AI is becoming the ‘new ZTE problem’

Artificial intelligence takes this strategic logic into an even more consequential realm.

If AI eventually becomes deeply embedded in manufacturing, finance, healthcare, education, government, scientific research and even national defence, dependence on proprietary foundation models controlled by a handful of American companies could recreate China’s earlier dependence on foreign chips and operating systems.

From Beijing’s perspective, this is a potential “second ZTE problem”.

Open-weight AI therefore carries particular strategic value. Open models can be downloaded, modified, fine-tuned and deployed on infrastructure under domestic control.

They cannot eliminate every dependency — advanced chips, semiconductor equipment, memory, electricity and specialised talent remain indispensable — but they can change the nature of those dependencies.

This also helps explain the increasingly divergent paths China and the US are taking in AI development.

The dominant US frontier model is:

Advanced chips → massive computing clusters → more powerful proprietary models

China, constrained by restrictions on access to advanced AI chips, has stronger incentives to pursue a different path:

Limited computing power → greater algorithmic efficiency → lower costs → open models → large-scale deployment

Chinese companies have consequently placed greater emphasis on model compression, mixture-of-experts architectures, inference optimisation, memory efficiency and hardware-software co-design.

This is where DeepSeek’s strategic significance lies. The important question is not whether any particular Chinese model has comprehensively surpassed the most advanced American models. What matters is that DeepSeek has demonstrated the viability of an alternative path for AI development: one based on less computing power, lower costs and greater openness.

Moonshot’s Kimi K3 further reinforces that possibility. Its performance on complex benchmarks has reached levels comparable with leading proprietary US models such as Claude and GPT, while retaining strong cost efficiency, further strengthening the competitiveness of China’s open-model ecosystem.

From technological independence to a ‘Chinese open AI ecosystem’

China’s strategy has now moved beyond the pursuit of technological self-reliance at home.

Open AI models can be adapted to the languages, cultures, regulatory systems and databases of different countries, and deployed on locally controlled infrastructure. Combined with Chinese telecommunications equipment, cloud computing, data centres and training programmes, they could form an entire technological ecosystem:

Connectivity → cloud → computing power → models → applications → training → standards

Such a model is particularly well suited to the Global South.

Many developing countries do not need the world’s absolute best AI model. What they need is AI that is good enough, affordable, customisable and, crucially, practical to deploy.

China’s future advantage may therefore lie not in building the world’s “smartest” AI, but in spreading increasingly powerful intelligence around the world at lower cost.

The central question in future US-China AI competition may consequently shift from who has the most powerful AI, to whose AI ecosystem the world will ultimately use.

The US may continue to enjoy frontier supremacy, while China competes for diffusion supremacy.

There must never be another ZTE

The most enduring legacy of the ZTE affair was Beijing’s realisation that industrial strength without control over critical technologies amounts to incomplete national power.

Economic growth remains important. But if that growth rests on critical foreign technologies beyond China’s control, the prosperity it creates is strategically vulnerable. Development and security have therefore become increasingly inseparable.

This explains why, even as China’s economy needs stronger macroeconomic support, Beijing continues to attach such importance to semiconductors, AI, robotics and advanced manufacturing.

Nor is China seeking complete self-sufficiency. Its objective is more complex — to remain connected to the global innovation system while gradually eliminating the ability of any country to exercise a decisive “technological veto” over China.

From the ZTE crisis to today’s drive for semiconductor self-reliance, advanced manufacturing and open AI, the same strategic memory runs through China’s policies.

In 2018, Xi Jinping discovered that China’s technological future could still depend on his having to make a personal appeal to an American president. He does not want to have to make such an appeal again.

That is the most enduring strategic legacy of the “ZTE shock”: the determination to build a China that can compete globally and remain connected to the world, but whose technology can never again be switched off at the touch of a button by Washington.

This article was first published in the Hong Kong Economic Journal as “中興通訊創傷 倒逼中國科技自立自強”.

Related: China tech firms drive AI data centre boom in Southeast Asia | Why hyper-efficient Singapore is Asia’s most AI-anxious society