TRUMP IS HELPING CHINA WIN THE AI RACE / PROJECT SYNDICATE

Trump Is Helping China Win the AI Race

As the US regulatory environment becomes increasingly chaotic, discouraging investment and innovation, China has streamlined regulations and fostered public-private partnerships. The implications for the future of American influence could be profound.

Karman Lucero

NEW HAVEN – In the aftermath of pandemic-era restrictions and a government crackdown on the tech sector, China’s AI companies spent much of 2023 in a state of malaise. 

While American firms surged ahead with increasingly powerful models and applications, Chinese regulators appeared intent on constraining domestic innovation through stringent and often ambiguous rules.

But Chinese AI companies managed to pull themselves out of this rut by adopting new strategies and reshaping institutional relationships between the authorities and private developers with a view toward streamlining domestic regulations. 

Since then, the landscape has shifted dramatically: the US AI industry now grapples with a chaotic regulatory environment under President Donald Trump’s administration, while China’s approach to AI governance has grown more flexible and supportive of innovation.

Several developments have contributed to this shift. 

For starters, the emergence of novel institutional mechanisms helped shape a more collaborative, multi-stakeholder approach to regulation, fueling the creation of new AI companies. 

One notable example is DeepSeek, China’s breakthrough AI firm, which grew out of a hedge fund. 

Others, like Zhipu AI and Moonshot AI – creator of the Kimi model – originated from a partnership between Tsinghua University and the Beijing Academy of AI, a nonprofit industry association established with support from municipal authorities.

These partnerships have gone beyond simply launching new AI developers – they have also infused China’s AI ecosystem with renewed energy and fresh talent. 

While many local governments continue to grapple with mounting fiscal pressures, especially in the wake of costly zero-COVID policies, leading cities have benefited from a combination of world-class infrastructure, targeted state subsidies for the tech sector (both direct and indirect), and synergies between hardware and software firms.

Closer collaboration between private companies and government entities such as the Ministry of Industry and Information Technology and the Cyberspace Administration of China (CAC) has influenced how regulations are applied, often in ways that benefit the tech industry. 

These agencies work through affiliated think tanks like the China Academy of Information and Communications Technology (CAICT) and industry groups such as the Cyber Security Association to convene regular dialogues with private-sector developers and academic experts.

Although not public, these discussions allow for a broader range of perspectives and have contributed to a regulatory environment that is more predictable and better aligned with the practical needs of tech innovators. 

Compared to the period before 2023, regulations governing AI developers have become more standardized and easier to navigate. 

The laws themselves have not changed, but what were once ambiguous or burdensome requirements now come with clearer guidelines for compliance.

In practice, regulatory scrutiny has moved away from the early stages of development, such as data collection and model training, to focus on specific applications. 

Most rules now target how AI models are deployed rather than how they are built, enabling companies to collect training data and develop foundation models relatively unimpeded. 

That said, the content generated by AI models remains heavily regulated.

Lastly, Chinese AI companies have invested heavily in developing innovative open-source models. 

While former US President Joe Biden’s administration considered restricting open-source AI development due to national-security concerns, ultimately deciding to support it, Chinese companies like DeepSeek and Alibaba forged ahead and released multiple open-source models now used worldwide.

The rapid rise of China’s AI sector stands in stark contrast to recent developments in the United States. 

Many AI companies initially welcomed the Trump administration’s deregulatory agenda, but the lack of detailed policy proposals, together with Trump’s vindictive approach to criticism, has created uncertainty and discouraged meaningful engagement. 

Moreover, sweeping cuts to federal research funding and severe immigration restrictions have choked the talent pipeline that has long fueled America’s AI sector, leading to growing unease across the industry.

America’s regulatory dysfunction has created room for Chinese AI companies to scale even faster. 

The rise of DeepSeek, China’s manufacturing and supply-chain dominance, and the perception of US self-sabotage have led to renewed confidence within the Chinese tech sector.

To be sure, this dynamic could change. 

The US Congress could still re-introduce the proposal for preempting state-level AI laws to create a more industry-friendly regulatory environment. 

The White House’s upcoming AI Action Plan, expected later this month, may also clarify the Trump administration’s approach to tech oversight.

At the same time, China’s confidence could become hubris. 

Historically, when the Communist Party believes it has gained a strategic edge, it often responds by tightening control. 

After decades of rapid growth and relative openness, for example, President Xi Jinping has ratcheted up political and regulatory pressure on private companies, especially in the tech sector.

Paradoxically, the more successful Chinese firms become, the more likely they are to draw government scrutiny. 

DeepSeek is a prime example: its rapid adoption – both at home and abroad – combined with limited content moderation, small compliance teams, and relatively lax censorship protocols could raise red flags for regulators like the CAC.

RedNote’s growing popularity in the US underscores the risks that can come with success. 

Within days of US TikTok users flocking to the platform, the CAC reportedly instructed RedNote’s government-relations team to separate them from Chinese users. 

While DeepSeek offers very different services, the institutionalized paranoia of the CAC and other regulators has proven to be indiscriminate. 

In fact, DeepSeek users have already noticed signs of increased censorship, with responses to “sensitive” questions reportedly changing in real time.

For now, however, as the Trump administration wages a Cultural Revolution-style assault on American industry and society, the Chinese government is doing everything it can to strengthen its position as a global technology leader. 

The implications for the future of American influence could be profound.

Karman Lucero is a fellow at the Paul Tsai China Center at Yale Law School.