China narrows AI gap with US — TimesLIVE
Chinese artificial intelligence developers are narrowing the gap with the United States in the capabilities of advanced models, despite US restrictions on China’s access to the most advanced chips and the equipment needed to manufacture them. Experts and reports, including from the Stanford Institute for Human-Centered Artificial Intelligence, indicate that China has become a counterweight to the United States in AI and will likely eliminate the US advantage this year, TimesLIVE reports.
Samm Sacks of the Johns Hopkins University Institute for America, China, and the Future of Global Affairs said that the gap between US and Chinese models is small and unstable.
Chinese models intensify competition
Beijing-based startup Moonshot AI drew attention with its Kimi K3 model, whose capabilities came close to those of US competitors. According to the Artificial Analysis platform, shortly after its release in July, the model ranked third in the world in intelligence. It is now in ninth place, behind new advanced models from Anthropic, OpenAI, Meta and SpaceXAI.
DeepSeek released its R1 model in January 2025. The article describes it as more economical than similar US systems. In April, the company presented a preview of the V4 model and announced improvements in knowledge, reasoning and agent capabilities — the ability to plan and carry out complex multistep tasks with limited human intervention.
More current news is available on the UA.News Telegram channel Telegram.
Startup Z.ai released GLM-5.2 in June and GLM-5.3 in August, with improved programming and agent-task execution capabilities. Alibaba introduced Qwen3.8-Max in July, calling it the most powerful model in its series.
Distillation allegations and AI risks
The FBI, the US National Security Agency and the US Cybersecurity and Infrastructure Security Agency said that Chinese AI developers allegedly obtained the capabilities of advanced US models through “aggressive, malicious and targeted” distillation. This involves training models to copy or imitate the functions of more powerful systems. US agencies also claim that Chinese companies may have used unauthorized access and other methods to extract restricted capabilities from US models.
China rejected the allegations, saying that distillation is used by many companies in the AI industry, and called on the United States not to make unfounded accusations. Lizzi Lee, a researcher at the Asia Society Policy Institute, noted that these claims remain disputed but have strategic significance.
China’s Minister of State Security Chen Yixin also warned about AI risks to the country’s political, ideological and cybersecurity. Among the possible threats, he named the mass spread of political rumors through deepfakes, data leaks, cyberattacks, and the use of AI to identify and exploit cybersecurity vulnerabilities.
