Why China's State Media Is Pushing to Replace English AI Terms

People’s Daily Calls for Chinese Equivalents to English AI Terms

People’s Daily, the official newspaper of the Chinese Communist Party, has called for English-language artificial intelligence terms to be dropped in favour of standardised Chinese equivalents. The commentary, published earlier this month, frames the move as part of a broader push to strengthen China’s “discourse power” in global technology governance, South China Morning Post reports.

The article comes at a moment when AI technologies have spread rapidly across Chinese industry, and English-language jargon has followed close behind. Terms such as “agent” — used to describe autonomous AI systems that can plan and execute multi-step tasks — and “LLM,” shorthand for large language model, have entered everyday Chinese tech vocabulary without translation, the commentary noted.

While the appeal is, on its surface, about clarity for ordinary Chinese readers, the deeper argument is strategic. People’s Daily contends that sustained reliance on foreign terminology would erode China’s ability to define the concepts that underpin the AI revolution — and, by extension, its authority to participate on equal terms in international standards-setting and policy debates.

The “Cognitive Dependency” Warning

Central to the commentary is the notion of “cognitive dependency.” People’s Daily warned that the uncritical adoption of English AI jargon would lead to a situation in which Chinese users, technology workers, and policymakers internalise a worldview shaped by foreign languages, ultimately undermining the country’s voice in global technology governance, South China Morning Post reports.

The newspaper elaborated on three specific risks flowing from this dependency. First, it argued that reliance on foreign jargon could deprive China of the authority to define key concepts in the AI field. Second, it suggested that the absence of an indigenous conceptual vocabulary could stunt the growth of original domestic theories. Third, it warned that the dynamic would hinder the development of a Chinese framework for talking about technology — a framework that could otherwise contribute to global debates rather than merely consume them.

Read together, the three risks describe a feedback loop. If Chinese technologists, regulators, and citizens describe AI systems using English concepts, they will tend to reason about those systems using English assumptions. Those assumptions, in turn, will shape the questions they ask, the problems they prioritise, and the solutions they propose. The commentary argues that breaking this loop requires deliberate intervention in the language itself — not waiting for organic, market-driven translation to occur.

The framing also reflects a recurring theme in Chinese policy discourse: technological sovereignty depends on more than hardware and software. It depends on the conceptual scaffolding through which a society understands and directs its tools.

Which English Terms Are in the Crosshairs?

People’s Daily’s commentary singled out specific terms that have already taken root in Chinese AI discourse. The two most prominent examples were “agent” and “LLM,” but the underlying target is broader. The article treats the propagation of English AI terms as a pattern, not a one-off borrowing, and warns that the pattern is accelerating.

“Agent” describes a class of AI systems designed to act autonomously on behalf of a user, often by chaining together multiple model calls, tool invocations, and decision points. The concept has gained traction in product roadmaps from major Chinese and American AI labs, and English-language marketing materials for agent platforms have become common in both markets. “LLM,” or large language model, refers to the foundational model architecture behind generative AI assistants and remains the dominant technical category for the most capable AI systems.

Beyond these two, the broader category of terms includes “prompt,” “fine-tuning,” “embedding,” “transformer,” “retrieval-augmented generation,” and dozens of other technical and product-oriented labels. Most of these have entered Chinese usage in raw form, even when descriptive Chinese equivalents exist or could be coined. The commentary’s argument is that this ad hoc borrowing will continue unless deliberate translation efforts are made.

People’s Daily noted that English AI terms are rarely self-explanatory to ordinary Chinese people, particularly when those terms lack direct parallels in everyday speech. The commentary implies that, without standardised Chinese replacements, the AI conversation risks becoming a domain that only English-speaking or English-fluent audiences can fully access — within China as well as internationally.

Why ‘Agent’ and ‘LLM’ Stand Out

Among the dozens of English terms entering Chinese usage, “agent” and “LLM” carry particular weight because they describe categories rather than features. A category term is the kind of word that organises an entire subfield: it explains what something is, what it does, and how it differs from other things. When a category term is borrowed wholesale, the conceptual framework that comes with it travels too. People’s Daily’s choice to highlight “agent” and “LLM” reflects this concern: these are not incidental jargon but load-bearing labels for the AI industry’s most important current products.

A History of Localizing Foreign Technology Terms

To ground its argument, the commentary drew a contrast with earlier waves of technology adoption in China, when foreign terms were deliberately translated into Chinese equivalents as part of broader efforts to localise new industries. The most prominent historical example cited is “dian nao” — literally “electric brain” — which became the standard Chinese term for computer.

The commentary’s framing treats that earlier wave of localisation as a deliberate policy and translation achievement, not an organic linguistic accident. The implicit message is that similar deliberation is required today if AI is to be integrated into Chinese society on Chinese terms.

The ‘Dian Nao’ Example

The “dian nao” precedent is instructive for how technological vocabulary can be deliberately shaped. When computers first entered widespread Chinese use, the English-language term “computer” was still relatively new in many parts of the world. Chinese translators chose a descriptive compound that captured the device’s function in native terms, prioritising clarity for ordinary users over fidelity to the English source. The term stuck: today, “dian nao” is the standard reference for personal computers across the Chinese-speaking world, and the original English borrowing has been largely supplanted in everyday speech.

Other precedents follow similar patterns. Compounds built from native Chinese roots have historically been preferred for foundational technologies, allowing users to grasp the category from the name itself. People’s Daily’s commentary implicitly asks whether AI will follow the same trajectory, with carefully crafted Chinese terms that capture both the function and cultural framing of the technology, or whether the field will be defined primarily by English-origin vocabulary that flows directly from American and European AI labs.

The asymmetry is notable. Until the early 21st century, key technology categories — radio, television, computer, internet — all received Chinese coinages that became standard. AI is the first major technology wave in which English terms have entered Chinese usage at scale before localised equivalents have been established. The commentary treats this as a problem to be addressed rather than a permanent state of affairs.

What “Discourse Power” Means in Global Tech Governance

The People’s Daily commentary invokes the concept of “discourse power,” a term that has gained prominence in Chinese policy discussions about the country’s role in international affairs. Discourse power refers to the capacity of a nation or civilisation to shape the conceptual frameworks through which global issues are understood, debated, and regulated.

In the AI context, the term carries practical weight. International standards bodies, regulatory sandboxes, and bilateral AI dialogues are increasingly establishing the technical and ethical vocabulary that governs how AI systems are deployed worldwide. The terms agreed upon in these forums often translate directly into national regulations, industry practices, and consumer protections. Whoever helps set the vocabulary helps set the agenda.

South China Morning Post reports that the People’s Daily commentary argued that, without a robust Chinese vocabulary for AI, China risks following rather than leading in these discussions. The implication is that the country should invest in formalised Chinese equivalents — ideally developed through coordination between linguists, technologists, and policymakers — that can serve as both domestic standards and internationally recognised references.

Implications for Chinese AI Companies

For Chinese AI companies, the commentary’s message carries operational implications. Domestic firms have built marketing, product, and developer-facing materials around borrowed English terms, partly because the technical literature remains predominantly English and partly because English terms have signalled global sophistication in product positioning. A shift toward Chinese equivalents could require substantial retooling of documentation, training materials, user interfaces, and product names.

It also affects cross-border communication. Chinese AI developers frequently publish research papers, technical blog posts, and conference presentations in English to reach international audiences — an audience that is expected to remain a meaningful constituency for Chinese AI firms seeking global relevance. Reconciling the push for Chinese-language standardisation with the need to participate in global research communities will be a practical challenge that companies will need to navigate carefully.

There is also a brand dimension. Some Chinese AI applications have built significant recognition around English-language product names and category labels. Translating or replacing these will require careful management of consumer expectations and search-engine discoverability, particularly for products that have international user bases.

Reactions and Industry Context

The commentary is part of a broader pattern of Chinese state media commentary on AI governance, ethics, and industry development. People’s Daily has periodically used its platform to signal policy direction, and the language about “discourse power” echoes themes that have appeared in Chinese academic and policy discussions about technology sovereignty for years. The state media’s attention to AI terminology is a continuation of longstanding sensitivities about linguistic influence, applied to a new technology.

Whether the commentary will translate into formal policy is unclear. People’s Daily’s editorial pages are influential, but commentaries are not binding regulations. The piece is, however, an unmistakable signal that the question of how AI is discussed — and in which language — has become a matter of strategic concern at the highest levels of China’s media apparatus.

For international observers, the commentary offers a window into how China is thinking about its position in the global AI landscape. The country is home to some of the world’s most aggressive AI deployment efforts, with major platforms serving hundreds of millions of users domestically. The vocabulary that emerges from this large-scale ecosystem will, in turn, shape how those systems are described, regulated, and understood both inside and outside China.

The longer-term question is whether Chinese-language AI terminology will develop the same kind of global reach that English-language terms currently enjoy. Early efforts to translate core concepts will likely serve as the foundation for that work, and the People’s Daily commentary is a clear signal that the work has begun. The depth of that effort — and the willingness of Chinese industry, academia, and government to coordinate on common standards — will determine whether the country’s “discourse power” in AI grows alongside its technical capabilities.

Why China’s State Media Is Pushing to Replace English AI Terms