China’s latest AI breakthrough signals a new phase in the technological rivalry with the United States
By Tom Harper
With the release of Moonshot’s Kimi K3 – one of the world’s most advanced AI models – all signs point to China becoming an AI superpower, marking a significant development in its wider competition with the United States.
This competition has not gone unnoticed in Washington. US Treasury Secretary Scott Bessent has called for further restrictions on Chinese AI firms over alleged intellectual property theft and access to restricted Nvidia chips. However, these concerns represent only part of the wider challenge posed by China’s growing AI sector.
Enjoying this article? Check out our related reads…
The growth of Chinese AI is the latest example of China’s broader technological transformation, demonstrating how Chinese firms have become capable of producing technologies once the preserve of advanced economies. It marks a dramatic shift from the Reform and Opening Up era of the 1980s and 1990s, when ‘Made in China’ was synonymous with cheap, poor-quality goods. With the development of Kimi K3 alongside earlier models such as DeepSeek, AI has become the latest field in which China is challenging the technological dominance of the West.
At the same time, the rise of Chinese AI highlights a different approach to development. While many leading Western AI companies have moved towards increasingly closed systems, Chinese firms have become strong advocates of open-source technology and open-weight models built around free or low-cost distribution.
This has helped drive the adoption of Chinese models across the Global South, where many organisations are priced out of using American alternatives. In doing so, China has also revived aspects of its long-standing ambition to present itself as a leader of the developing world. The accessibility of Chinese AI models is therefore likely to strengthen Beijing’s influence while posing a growing challenge to the dominance of established Western systems, as demonstrated by the rapid success of DeepSeek.
Chinese AI models have also found a receptive audience in developed economies despite the political controversy surrounding them. Companies including DoorDash, Siemens and Airbnb have all adopted AI tools developed in China. Their appeal lies in a combination of lower costs and performance that is increasingly comparable with American rivals. More recently, a Chinese model was reportedly used by New York-based AI company Hugging Face to help defend its systems after a rogue OpenAI model gained access through stolen credentials. Whether for cost, performance, or resilience, Chinese AI is proving capable of competing in areas once dominated by American and European firms.
This progress is perhaps best illustrated by Moonshot’s Kimi K3 model, which the company describes as one of the world’s largest open-weight AI models. Although its overall performance is reported to sit just behind Anthropic’s Claude, it has nevertheless achieved what analysts describe as frontier-level performance. Analysts at Bank of America have also praised the model, noting that Moonshot has made significant advances despite lacking access to the most advanced semiconductor chips. In many ways, Kimi K3 has become the poster child for the challenge posed by Chinese AI.
As Chinese AI continues to advance, technological competition between China and the US is likely to intensify. Calls in Washington to further restrict Chinese AI development, including allegations that Kimi benefited from access to restricted Nvidia chips, underline the increasingly strategic nature of the rivalry. At the same time, the US is pursuing initiatives designed to reduce allies’ dependence on Chinese semiconductor supply chains and counter Beijing’s growing influence in AI. Should these efforts gather momentum, the world could find itself divided between competing AI ecosystems, with governments and businesses increasingly forced to choose between American and Chinese models.
The defining feature of China’s AI strategy is its dual purpose: achieving technological parity at home while expanding diplomatic influence abroad. By undercutting US pricing and promoting a non-Western model of AI governance, China is steadily building a network of AI-dependent partners. The result may not be a US-led monopoly on artificial intelligence, but an increasingly contested duopoly in which influence is determined as much by accessibility and affordability as by technological innovation.
