Source : INDIA TODAY NEWS
The United States and China are continuing to race to build more powerful artificial intelligence models even as both sides voice concerns about the risks of the technology. The US has remained ahead for years in AI model capabilities, research and citations, but experts say China is rapidly narrowing the gap despite US-led restrictions on advanced chips and the machines needed to make them.
advertisement
A recent report by the Stanford Institute for Human-Centered Artificial Intelligence said China has emerged as an AI counterweight to the US and this year appears to be erasing the American lead. The debate has gained more significance ahead of Chinese President Xi Jinping’s expected visit to Washington this month to meet US President Donald Trump, with experts saying the future of AI will depend heavily on what the two countries do and do not do. “The gap between US and Chinese models is narrow and fragile,” said Samm Sacks, a senior fellow at the Johns Hopkins School of Advanced International Studies’ Institute for America, China, and the Future of Global Affairs.
Among Chinese companies, Beijing-based startup Moonshot AI drew wide attention with Kimi K3, which came close to US rivals in capabilities. Artificial Analysis ranked it No. 3 globally in model intelligence in July, shortly after its release, behind Anthropic’s Claude Fable 5 and OpenAI’s GPT-5.6 Sol. Kimi K3 has since slipped to ninth place, behind newer frontier models from Anthropic, OpenAI, Meta and SpaceXAI, underlining how quickly the rankings are changing.
DeepSeek was the first Chinese startup to shake global stock markets by showing that Chinese AI models could challenge the US lead. Its R1 model, released in January 2025, was cheaper than similar US models. In April, the company previewed its V4 model and said it brought major gains in knowledge, reasoning and “agentic” capabilities, referring to the ability to plan and carry out multiple complex steps with limited human guidance. Startup Z.ai rolled out GLM-5.2 in June and drew comparisons with the most advanced models from Anthropic and OpenAI. It followed that with GLM-5.3 in August, with stronger coding and agentic capabilities that lifted it in global rankings. In July, Alibaba previewed its Qwen3.8-Max model, which it said was the most powerful in the series and a challenger to Anthropic’s Claude Fable and OpenAI’s GPT.
At the same time, China has publicly raised concerns about the risks AI could pose to security and governance. In an article published on Sunday, state security minister Chen Yixin said the technology could create threats ranging from political and ideological security risks to cyberattacks, data leaks and pressure on governance. He wrote that AI could threaten China’s political, institutional and ideological security if used by “hostile forces and individuals with ulterior motives”. He also warned that some people could use AI-generated deepfake audio and video to create “political rumors” on a massive scale.
Chen specifically cited the capabilities of Anthropic’s Claude and OpenAI’s GPT as examples of how AI could make it easier to identify and “weaponize” cybersecurity vulnerabilities. He said certain nations and groups can now “execute complex hacking tasks”. Even as top Chinese leaders have called for faster AI development and technological self-reliance, they have also stressed the need to control the risks and build a global framework for AI governance.
In Washington, the focus has been on whether Chinese developers have copied or extracted capabilities from American models. The US has accused Chinese AI firms of what it calls “industrial-scale distillation”, where models are trained to copy or imitate more powerful systems. In a report issued on September 8, the FBI, the National Security Agency and the Cybersecurity and Infrastructure Security Agency said Chinese AI developers extracted capabilities from some of the most advanced US models in “aggressive, malicious, and targeted distillation activities”. On the same day, US Treasury Secretary Scott Bessent said at an event, “the Chinese distill our models and they can never get ahead of us.” China rejected the allegations, said distillation is widely used across the industry, asked the US to avoid making “unfounded accusations”, and accused Washington of pursuing a “monopoly of the AI industry”.
Lizzi C Lee, a fellow at the Asia Society Policy Institute’s Centre for China Analysis, said distillation is a standard technique across the AI industry, but the US agencies had made the more specific allegation that Chinese firms used unauthorised access and other methods to extract restricted capabilities from American frontier models. “These are still contested allegations at this point but I think the strategic implication is already important,” she said. Lee added that the United States had spent significant effort controlling inputs into frontier AI, especially chips, but now “it seems that it is finally discovering that it may also have to think about the outputs of frontier AI as a form of technology transfer”. She said, “The US labs spend enormous amounts of compute pushing out the frontier. But Chinese developers can learn from those systems,” and added, “And Chinese companies then (can) release very capable open-weight models that spread globally.” Overall, the AI contest remains centred on China’s fast-growing model ecosystem, Beijing’s worries about security risks, and Washington’s claims that Chinese firms are drawing on American advances even as the two sides compete for influence in the technology.
With PTI Inputs
– Ends
SOURCE :- TIMES OF INDIA
