While the U.S. is the "superhero" of global AI, China is the "supervillain." This claim was made by U.S. House Foreign Affairs Committee Chairman Brian Mast at a recent think tank event on "Securing America's AI Advantage: A Discussion on U.S. Export Control Policy."

Just two days before the event, U.S. Treasury Secretary Scott Bessent said China surpassing the U.S. in AI was the "biggest risk," outweighing concerns over safety and job losses.

Such rhetoric highlights a troubling reality: For some American politicians, AI has become deeply politicized. China's rapid progress in AI is increasingly portrayed as the foremost challenge to U.S. technological dominance. Statements like those made by Mast and Bessent are intended to demonize China's AI development and provide political justification for further U.S. efforts to contain and suppress China's technological progress.

This narrative, however, is convincing only to themselves and a handful of those who are either misinformed or deliberately obtuse. China's AI development has never posed a threat to the world. On the contrary, it is generating tangible benefits for countries across the globe — including the U.S.

Due to the sweeping U.S. export controls on advanced chips, Chinese AI companies have continued to pursue independent innovation while embracing an open-source development model. In recent months, they have successively released open-source large language models such as DeepSeek-V4, Zhipu AI's GLM-5.2, and Meituan's LongCat-2.0. These models not only contribute new technical approaches to global AI research but also provide affordable, accessible solutions for countries — particularly developing economies — seeking to accelerate digital and intelligent transformation.

A Foreign Policy article in June noted that developers across Africa are widely adopting Chinese AI platforms, including DeepSeek, Qwen, and Kimi, to build AI models capable of supporting local languages. According to the article, Chinese models have become the "overwhelming choice" for African developers.

China's AI models are openly available to users worldwide, without discrimination based on nationality. Statistics from OpenRouter, a major AI model aggregation platform, indicate that more than 60 percent of the platform's token traffic is now processed by Chinese models. Reports also indicate that major U.S. companies, including Microsoft, Airbnb, and Coinbase, are exploring ways to integrate Chinese AI models into their businesses.

By contrast, leading U.S. AI companies such as OpenAI and Anthropic continue to impose strict restrictions on users from China. Anthropic also faces allegations that its AI coding tools contain mechanisms specifically designed to monitor or restrict Chinese users.

Equally noteworthy is Washington's growing tendency to invoke "national security" to limit international access to its most advanced AI models. In reality, this reflects an "America First" approach to AI policy — one aimed at preserving U.S. technological supremacy by restricting other countries' access to cutting-edge models and widening the global technology gap.

On June 12, the U.S. Department of Commerce reportedly imposed export controls on two Anthropic AI models, temporarily restricting access for foreign organizations and individuals. Later that month, OpenAI's GPT-5.6 model was reportedly released under government oversight, with access limited to government-approved "trusted partners."

China is not only one of the world's leading developers of AI technologies but also a major advocate of global AI governance. In October 2023, China introduced the Global AI Governance Initiative, becoming one of the first major countries to propose a comprehensive international framework for AI governance.

The initiative emphasizes AI for good, safety, fairness, and inclusiveness, while specifically calling on all countries to "adopt a prudent and responsible attitude to the research, development, and application of AI technologies in the military field."

The U.S., however, appears to be moving in the opposite direction by accelerating the "militarization" of frontier AI technologies.

Last month, the Financial Times reported that the U.S. National Security Agency had partnered with Anthropic to employ its Mythos AI model in offensive cyber missions. Shortly afterward, U.S. media outlets reported that the U.S. military had begun using xAI's Grok Gov model to support lethal military operations, claiming the system had assisted in striking approximately 2,000 targets within a 96-hour period.

As the world's two leading AI powers, China and the U.S. share a responsibility to advance the intelligent transformation of human society, strengthen international cooperation on AI governance, and jointly manage the security risks posed by rapidly evolving technologies.

What the U.S. should do is not continue to smear and suppress China's AI development, but rather reflect on its own role in shaping the future of global AI. It cannot proclaim itself the world's "superhero" while acting increasingly like the very "supervillain" it claims to oppose.