Open AI models fuel new battle over global technology dominance

Chinese-developed artificial intelligence models are reshaping the global AI landscape, challenging long-held assumptions about U.S. technological dominance and intensifying debate over how advanced AI should be regulated.

Open and open-weight models developed by Chinese companies, including DeepSeek, GLM-5.2 and Kimi K3, have gained traction among developers by lowering the cost of AI development and allowing greater flexibility for research and experimentation.

The shift comes as major U.S. technology companies, including Nvidia, Microsoft, Meta and IBM, have urged U.S. lawmakers to avoid imposing broad restrictions on open AI models, arguing that overly restrictive rules could weaken U.S. competitiveness and push innovation overseas.

Their intervention coincides with growing concern in Washington over China’s rapid advances in AI and discussions about tighter controls on advanced AI systems.

At the same time, Anthropic Chief Executive Dario Amodei has called for stronger government oversight of advanced AI, arguing that increasingly capable systems require robust safety measures. Anthropic has also backed tighter export controls on advanced AI chips and restricted access to some of its models for China-linked entities.

The contrasting positions have highlighted a broader debate over whether proposals to regulate advanced AI are driven primarily by legitimate safety concerns or increasingly influenced by geopolitical competition.

Experts broadly agree that powerful AI systems pose real risks, including the spread of misinformation, cyber threats, fraud and the potential misuse of autonomous technologies.

Governments have said regulation is needed to address such risks while supporting innovation.
Critics, however, argue that regulations should be applied consistently and warn against measures that could disproportionately disadvantage emerging competitors or restrict open research.

The debate has also exposed divisions within the U.S. technology industry over the future direction of AI development, with some companies advocating for open models that are widely accessible to developers and researchers, while others favour tighter controls over increasingly powerful systems.

Analysts say China’s progress in open AI has become a competitive challenge rather than simply a technological milestone, prompting governments and companies to reassess assumptions about long-term leadership in artificial intelligence.

The outcome could shape whether the next phase of AI competition is driven primarily by technological innovation and market adoption or increasingly by government policy and regulatory frameworks.