China Projected to Lead in LLMs by 2030

Editor's NoteChina's artificial intelligence (AI) technology is threatening South Korea's industrial ecosystem and national security. Backed by massive capital and infrastructure, Chinese companies are using cost-effectiveness and open source as weapons to encroach on sectors ranging from tech to the public and financial sectors. This is no longer just a matter of adopting technology; the risk to software (SW) supply chain security is becoming a reality, with data control and national sovereignty potentially slipping overseas from the design stage. This article examines the technological onslaught of Chinese AI and explores strategies for the survival of South Korea's AI industry, including securing an "independent AI foundation model" to protect national data sovereignty.

"U.S. export controls on advanced graphics processing units (GPUs) for China did slow China's pursuit of the United States, but they also gave rise to a uniquely Chinese AI ecosystem built around computational efficiency and open access."

"They Blocked China's Chips, and It Became a 'Value-for-Money Monster'..." China's Counterattack, Copying the U.S. and Selling It for Peanuts [China Shock, K-Industry on the Brink | AI] 1 View original image

As China rapidly takes over global markets with so-called "cost-effective AI models," a new forecast says it will lead the world in large language models (LLMs), a core AI technology, by 2030. While U.S. frontier AI companies operate on a closed model, offering their models only through application programming interfaces (APIs), China is aggressively expanding its reach by using open-weight models, which make model weights available, to broaden its user base.


On October 28, Bloomberg Intelligence (BI) said in a report that China would lead the world in eight of 17 key industries, including LLMs, by 2030. "China will lead the world in LLMs by 2030," the report said, adding, "The United States and China are likely to share leadership in AI technology, and neither side is likely to gain an overwhelming advantage." (Related article: "China to Lead the World in AI Language Models by 2030, Sharing Tech Supremacy with the U.S.")


China is expanding its developer base with models that emphasize low costs and openness. Jung Junhwa, a legislative researcher at the National Assembly Research Service, explained, "China, which has found it difficult to secure large-scale computing resources due to the U.S. government's export controls on advanced GPUs, has concluded that it cannot catch up with frontrunners through a closed subscription model. Instead, it has chosen a circular approach: securing global users and data at low prices, then reinvesting them in improving its own AI technology." He added, "China's approach is unlikely to shift away from openness. Instead, it will converge on a 'dual structure' in which general-purpose models are distributed as open-weight models while only the most advanced models are controlled."

U.S. Export Controls Instead Gave Rise to a Chinese Ecosystem Focused on Computational Efficiency

As the U.S.-China competition for AI supremacy heats up, companies are also stepping up their efforts to counter one another. On October 12, Anthropic announced it would actively respond to Chinese AI companies' attempts to replicate its model capabilities and take defensive measures. Anthropic criticized the companies, saying, "Chinese AI companies are attempting 'distillation attacks,' using large numbers of accounts and automated requests to replicate the performance of the AI model Claude." A distillation attack involves making a large number of queries to a competitor's high-performance AI model, collecting its responses, and using them to train a subsidiary model.

"They Blocked China's Chips, and It Became a 'Value-for-Money Monster'..." China's Counterattack, Copying the U.S. and Selling It for Peanuts [China Shock, K-Industry on the Brink | AI] 1 View original image

Anthropic claimed that Chinese AI research institutions had tried to collect large volumes of Claude's responses and use them to train competing models. It said customers accounting for approximately 5% to 7% of its revenue from January to September this year were among those blocked. In February, Anthropic had also announced that DeepSeek, Moonshot, MiniMax and others had generated more than 16 million interactions through about 24,000 fake accounts. Anthropic has since restricted proxy access, blocked accounts, and introduced new technical measures.


As AI emerges as a key strategic asset directly tied to national survival, not just technology, companies are tightening access controls to protect the core technology behind their AI models. Conflicts are also intensifying between openness and technology protection in AI services, and between customer acquisition and revenue growth. As China has gained a competitive edge in AI models in a short time and at low cost, the United States has stepped up its efforts to contain China. The highest-performing AI models are not immediately made available to the public; instead, they are first distributed on a limited basis to select institutions vetted by the government.


Hong Gwanhui, chief information security officer (CISO) at LG Uplus, said, "The U.S. government became involved in selecting partners for Project Glasswing for both Anthropic's Mythos Preview and Mythos 5, bringing them under national control." He added, "The deployment of advanced AI models is expanding beyond simple business considerations between companies into the realm of political and diplomatic relations between governments and companies, and between countries."


"They Blocked China's Chips, and It Became a 'Value-for-Money Monster'..." China's Counterattack, Copying the U.S. and Selling It for Peanuts [China Shock, K-Industry on the Brink | AI] 1 View original image

Concerns Grow Over Supply Risks for Domestic Companies Using Chinese Open-Weight Models

The problem is that Chinese AI is making broad inroads into the South Korean market with free or nearly free, cost-effective open models. In South Korea, neither U.S. nor Chinese models are guaranteed to be supplied reliably, and domestic companies that have used Chinese open-weight models for their cost-effectiveness may face new supply risks. At a recent parliamentary audit, controversy erupted over the Korea Investment Corporation (KIC), which manages assets worth about 339 trillion won, using a language model from China's Alibaba and a search model from a Chinese research institute on the U.S. government's export-control list in its in-house AI system.


As the use of Chinese AI models grows in both private companies and the public sector, the government is pushing for a transition to domestically developed models. Bae Kyunghoon, Deputy Prime Minister and Minister of Science and ICT, said, "The reality is that many industries and public institutions are using Chinese AI models." He added, "Once the results of the third evaluation of independent AI foundation models are released, we will further expand the use of domestically developed AI models."


According to South Korea's developer community, the use of Chinese models is widespread. Lim Junghwan, CEO of Motif Technologies, explained, "Chinese AI companies such as DeepSeek, Moonshot AI, and Alibaba make the weights of their fully trained models available on platforms such as Hugging Face, allowing anyone to run them on their own servers and modify them for specific purposes. As a result, their reach is expanding, particularly among developers."



A developer who asked not to be named said, "Users are highly satisfied with the performance of Anthropic's Claude Fable 5, but it is too expensive." The developer added, "The price of Kimi K3, released by China's Moonshot AI, is up to 70% lower than Anthropic's top models, so a considerable number of South Korean companies are using Chinese AI. However, we are keeping a close eye on recent moves by Chinese AI companies to abandon their low-price strategy."


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing