The U.S. artificial intelligence (AI) industry and policymakers are closely watching the rapid rise of Chinese AI models.

Major U.S. AI companies such as OpenAI and Anthropic are warning that China's low-cost, high-performance AI models could pose threats to national security and industrial competitiveness, but some critics argue that this is a move to check competitors in the market.

As Chinese companies have recently disclosed open AI models that have reached levels threatening the performance of U.S. companies, analysts say the AI supremacy competition has entered a new phase.

▲ Chinese Open AI Models Surging…U.S. AI Industry on Alert

According to the Wall Street Journal (WSJ) on the 20th (local time), Silicon Valley and Washington political circles have begun full-fledged debate over whether U.S. companies should be allowed to freely use Chinese AI models.

At the center of the controversy are recently unveiled "Kimi K3" from Chinese Moonshot AI and "Qwen 3.8 Max" from Alibaba.

Both models were released in an "Open Weight" format that allows users to freely download and modify them to suit corporate data, and in some performance evaluations, they have demonstrated competitiveness on par with the top U.S. AI models.

With positive reviews from investors and users continuing, assessments indicate that there are significant changes appearing in the competitive landscape of the AI market.

▲ OpenAI: "AI Ecosystem Will Collapse Without Regulation"

OpenAI has argued that the spread of open AI models could threaten the long-term sustainability of the AI industry.

Dean Ball, OpenAI's head of strategic futures, argued that "if open AI dominates the market, 'AI communism' where AI becomes a form of public good provided by the state could become reality."

He described this situation as a "dystopian future" and predicted that the U.S. government might pursue policies to reduce the use of open-source AI models.

However, he later clarified that his remarks did not mean that Chinese AI itself should be banned.

▲ "If Free AI Spreads, Investment in Advanced AI Will Become Difficult"

OpenAI has also pointed to the investment structure of the AI industry as another problem.

Currently, cutting-edge AI companies are investing billions of dollars in developing ultra-large AI models and must continuously secure massive computing resources.

However, concerns have been raised that if most users use only low-cost open AI models with minimal payment, it may become difficult to secure investment funds for next-generation AI research and development.

OpenAI noted, however, that Dean Ball's remarks represent his personal views and not the company's official position. CEO Sam Altman has also acknowledged in the past that the strategy of developing only closed models was not appropriate, and OpenAI currently is developing its own open models.

▲ Anthropic: "Powerful Open AI Is a Security Threat"

Anthropic also warned that open AI models could pose serious security risks.

CEO Dario Amodei expressed concern that if AI models with powerful cyber-attack capabilities become freely downloadable to anyone, the potential for misuse could be very large.

In a recent interview, he emphasized this issue as "a very serious concern."

▲ Criticism Also Spreading: "Regulatory Demands to Check Competitors"

Meanwhile, AI industry researchers and some investors have suggested that the regulatory demands of U.S. AI companies are a strategy to check competitors.

Since OpenAI and Anthropic are preparing for future initial public offerings (IPOs), an analysis has emerged that their company value could be affected if low-cost Chinese AI models capture market share.

In particular, Chinese AI models are being provided at much lower costs than U.S. products, and the number of adoption cases by U.S. companies is rapidly increasing.

Consequently, among investors, questions are growing about whether OpenAI and Anthropic can continue to maintain their current market dominance.

[AP/Yonhap News]

▲ Trump Administration Divided on Regulation

It is known that opinions are divided within the U.S. government regarding regulation of Chinese AI.

Security authorities reportedly considered including Chinese AI companies in trade sanctions or strengthening warning measures, and even reviewed executive orders targeting open AI models.

However, due to disagreements within the administration, no specific regulations have been implemented to date.

The "advance submission requirement for cutting-edge AI models" included in recently announced executive orders is also reportedly experiencing implementation delays.

▲ Political Conflict Over AI Regulation Intensifies

David Sacks, who serves as a White House AI policy advisor, criticized OpenAI's arguments as "a strategy to exclude competitors through regulation."

He pointed out that using regulatory uncertainty as a competitive tool is an unacceptable practice.

Meanwhile, OpenAI, Anthropic, Google DeepMind, and others maintain the position that government oversight becomes more necessary as AI technology becomes more sophisticated.

Experts assessed that inexpensive yet high-performance open AI models are likely to spread rapidly across global industries.

Austin Carson, CEO of SeedAI, analyzed that "while AI is now a technology directly linked to national competitiveness, the importance of affordable, open AI models that companies can directly control will continue to grow, and in the open-source ecosystem, one cannot achieve victory through exclusion."