China Rejects AI CEOs' Call for Moratorium, Calls It 'Fear Mongering'

China's government has dismissed calls from a group of prominent artificial-intelligence executives to slow down development of advanced AI systems, describing the appeal as "fear mongering" rather than a serious safety concern. The statement, reported by CNBC, marks one of the sharpest public rebukes yet from Beijing toward the Western executives who have spent the past year warning about existential risks from frontier models. The dispute puts the world's two largest AI powers on opposite sides of a debate over whether the technology's dangers justify hitting the brakes.

The pushback follows an open letter signed by hundreds of industry figures, including several high-profile CEOs, urging labs to pause training runs more powerful than GPT-4 for at least six months. In China, officials and state-linked commentators have argued that such a pause would freeze the current competitive order in place and hand incumbents like OpenAI, Google and Anthropic time to consolidate their lead. The Chinese response frames the safety argument as a commercial maneuver dressed up in ethical language.

China's own AI sector has moved quickly despite earlier regulatory tightening. Baidu, Alibaba, Tencent and a cluster of well-funded startups have released large language models at a steady clip, and Beijing finalized rules last year governing generative AI services while encouraging domestic development. That posture contrasts with the European Union's slower, risk-tiered AI Act and with the voluntary commitments several U.S. labs signed at the White House. For Chinese firms, a global moratorium would arrive at the moment they are trying to close a gap in compute and model quality.

The exchange also exposes how thin the consensus on AI safety has become. Some researchers argue a pause is unenforceable because it depends on voluntary compliance from actors who may not share the same incentives. Others counter that dismissing safety concerns outright risks normalizing a race with little oversight. Neither side has offered a mechanism that would bind labs across borders, which is precisely why the letter's signatories framed their request as a starting point rather than a binding treaty.

What happens next will hinge on whether the moratorium idea gains traction in Washington or Brussels, where it could translate into licensing requirements or compute thresholds. China's rejection signals that any global pause framework will struggle without buy-in from the country building some of the fastest-moving models. The open question is whether safety talks can continue at all when the two leading AI blocs treat the same proposal as either prudent caution or a competitive tactic.