China Balances AI Safety Pushback With Rapid Development

China is pushing back against US warnings over rapid AI development, even as it advances its own safety framework. Education entrepreneur Boris But had a "penny-drop moment" about two years ago when clients began citing chatbots instead of rival firms, prompting him to develop his own AI study tool.

The pace of AI development in China is second only to the US, with analysts saying frontier American models remain ahead. Anthropic CEO Dario Amodei recently called for a "worldwide pacing of the frontier" requiring cooperation with China. Omdia's Lian Jye Su said China sees such calls as a bid to lock in US advantage, viewing AI as a "core sovereignty capability."

China's foreign ministry responded that "fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance." This week China released its third AI safety governance framework, covering agent risks, cybersecurity threats and recursive self-improvement. Researcher Gabriel Wagner said there is more common ground between Amodei and Beijing than assumed.

President Trump, who this week called fears AI could wipe out humanity a "hoax," is set to meet Xi Jinping in Washington next week with AI on the agenda. Analysts are sceptical of a breakthrough. China has had an algorithm registry since 2022, and warnings this year suggest a shift "from controlling what AI says to controlling what it does." National regulations now aim to curb emotional dependency on companion AIs, while AI is encouraged across the economy in the latest Five Year Plan.

Carnegie's Leia Wang said Chinese companies do less voluntary monitoring than US ones, and neither government seems prepared for a loss-of-control incident, though China's regulation is "iterative." Black Lake Technologies said safety concerns in manufacturing are "concrete and operational," with reliability mattering most. Digital health service A-fu recently hit 150 million users. Ni Tao of yangtzeer.com said most people treat AI as a useful everyday tool, with anxiety about jobs coexisting with a belief China must master AI rather than slow down.

Source: The Guardian