AI regulation could boost US security and slow China, analysis says

Donald Trump has opposed tighter AI rules, warning they would hand China a strategic edge. A new analysis says stronger US regulation could improve security and slow Beijing's progress.

In Short

• Polling shows most Americans across parties favour stronger oversight of AI
• Anthropic reported foreign actors misusing Claude for fraud, surveillance and weapons
• Google flagged attempts to exploit Gemini for hacking, espionage and disinformation

US President Donald Trump has rejected calls for tighter regulation of artificial intelligence, arguing that slowing the technology would hurt American interests and allow China to gain an edge. In recent remarks, he said any move to slow AI progress would threaten the United States in a strategic race with Beijing.

But a new analysis argues the opposite case: that stronger regulation of the US AI industry could improve national security and also slow China's progress. It says unregulated AI poses risks ranging from cyberattacks and weapons development to threats to critical infrastructure, while China's advances are closely tied to the pace of progress in the US.

Trump has repeatedly argued against strict oversight of AI. "Whoever wins AI, WINS!" he wrote on social media on September 21, 2026, while naming China as the main threat to America "winning" AI. He later told the United Nations that his administration rejected "any attempt to construct a globalist scheme to control" artificial intelligence.

The analysis said concerns over unchecked AI growth are shared by leading industry and political figures across the world. It cited polling that found about eight in 10 Americans support more AI regulation, including majorities among Democrats, Independents and Republicans. The risks listed include foreign states or terrorist groups using AI to build more lethal conventional or mass-destruction weapons, as well as increasingly sophisticated cyberattacks that exploit security weaknesses.

It also pointed to recent disclosures by technology companies. On September 10, 2026, Anthropic said foreign actors, including some based in China, Russia and Iran, had used Claude in cyber, influence, surveillance and fraud operations and to support weapons development. The report also described "five case studies of actors using our models in ways that could support biological weapons development". Google, too, said actors linked to China, Russia, North Korea and Iran had tried to use Gemini for hacking, disinformation, espionage and criminal activity.

On Trump's argument that the US must move faster than China, the analysis said American firms still lead the world in advanced models, chips, computing power and investment, even though Chinese models are not far behind. It said China has used a method known as distillation to train its own models on advanced US systems, producing models that are nearly as powerful but use fewer resources. Matt Sheehan, a senior fellow at the Carnegie Endowment for International Peace, said Beijing's reliance on distillation stems from its limited computing power, estimated at about one-tenth of that of the US. The analysis compared the relationship to a speedboat pulling a water skier: if the US slows down, China would also be forced to slow.

The article also said the case against US regulation assumes China has no similar concerns about uncontrolled AI growth, which it called incorrect. It noted that Beijing introduced a major set of AI rules in 2021 and that Chinese officials have recently called for greater domestic and international oversight. Xi Jinping, while backing AI development, called for strong regulation and oversight at the 2026 World AI Conference in Shanghai. At a UN Security Council meeting on AI and international security on September 23, 2026, China's permanent representative to the UN, Fu Cong, called for "a consensus-based global governance framework" for AI to "build institutional guardrails that preserve international peace and security". In sum, the analysis said tighter US regulation could both address the security risks of unchecked AI and slow China's access to cutting-edge advances.

With PTI Inputs

- Ends