Artificial intelligence appears to be advancing faster than governments can agree on how to control it.

The risks are becoming harder to ignore. AI systems are being used to carry out increasingly complex tasks, while concerns are growing over job losses, cyberattacks and the possibility that future systems could start operating on their own will. 

There have even been publicly expressed fears by insider experts that future superintelligent AI systems could lead to humanity's extinction within a decade. 

Yet the world's two leading AI powers, China and the United States, are taking notably different approaches to the question of how much regulation is enough.

In Washington, some technology leaders are calling for stronger safeguards, while President Donald Trump has dismissed warnings about AI posing an existential threat as a “hoax.”

In Beijing, policymakers have taken a different approach: accelerating AI development while building rules specifically designed to prevent systems from escaping human control.

China's approach: develop AI, but keep it governable

Policymakers in China have generally framed the risks as serious, but manageable. Through regulation and human oversight, Beijing believes the power of AI can be properly harnessed. 

A national AI safety framework released in 2024 explicitly considered scenarios in which future AI could autonomously obtain external resources, replicate itself, develop self-awareness and seek greater power.

An expanded framework published in 2025 warned that AI could potentially experience a sudden and unexpected leap in intelligence before acquiring resources, replicating itself and seeking power.

It introduced the principle of “trusted application, preventing loss of control”.

The message is clear: AI development can continue, but systems must remain subject to human control.

China is also working on a mandatory national standard for AI agent safety.

But none of this means China is advocating a slowdown in AI development. In fact, quite the opposite is true. 

Beijing has made the integration of AI across industries a major economic priority, viewing the technology as a potential driver of productivity and growth.

Washington faces a different debate

In the US, the argument has become increasingly political.

AI executives including Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have called for stronger oversight, warning that the rapid development of increasingly powerful systems could create risks that are difficult to contain.

While industry leaders have recognised that AI isn't going anywhere - with Sam Altman claiming that calls to halt development are akin to ditching electricity - they've firmly placed the regulatory ball in the government's court.  

But Trump has rejected those warnings, arguing that excessive regulation could undermine US global competitiveness.

He has also defended the rapid construction of data centres needed to power AI systems, describing opposition to AI and data-centre expansion as part of a “SICK conspiracy”.

It appears as though Trump is at odds with some of the technology industry's most prominent figures - a position less than ideal for a booming sector. 

The risks are already here

The arguments over AI aren't simply about hypothetical superintelligence.

Data centres require huge amounts of electricity and water. Automation could transform employment in industries ranging from administration to software development. AI-generated misinformation and cyberattacks are becoming increasingly sophisticated.

Microsoft co-founder Bill Gates has warned that AI could either become “the greatest equalizer” or a major source of injustice. His foundation has pledged $1 billion over two years to expand access to AI tools for healthcare, education, local-language models and agriculture.

For China, the challenge is to expand AI's economic and social benefits while ensuring that increasingly capable systems remain controllable.

For the US, the debate is increasingly about how to maintain technological leadership without allowing regulation to become a competitive disadvantage.

But artificial intelligence continues to develop at breakneck speed, and whether tomorrow's world could arrive before governments have had the chance to map it out, remains to be seen.