Chinese AI Models Gaining Traction in the US
Artifical intelligence from China is gaining popularity in the United States. Raffi Krikorian, chief technology officer at Mozilla, recently switched to using Chinese startup Moonshot’s AI, Kimi K3, for his daily activities. He found it more responsive compared to the American Anthropic’s Claude Fable chatbot.
This shift is part of a larger trend as Americans increasingly choose Chinese AI systems, known for their affordability and efficiency. Companies like Coinbase are adopting these systems to reduce costs, causing concern among U.S. tech giants, but a ban seems unlikely.
U.S. restrictions prevent China from accessing advanced technologies such as AI chips, and further sanctions might be imposed to protect American intellectual property. Moonshot has been accused by the Trump administration of developing Kimi K3 using methods that are questionable yet not illegal. These claims have stirred accusations of intellectual property theft, denied by Beijing.
Growing Affordability and Usability of Chinese AI
China’s AI models offer a cost-effective alternative to American models. Startups like Z.ai and Moonshot have launched intelligent models that compete with those from OpenAI and Anthropic. Alibaba showcased its Qwen3.8 Max AI model, and DeepSeek introduced V4, challenging OpenAI’s GPT and Google’s Gemini.
These developments appeal to users searching for affordable solutions. Curt Meinhold, a tech executive, prefers Chinese models for tasks where top-tier AI capabilities are not necessary. He emphasizes that Chinese models are sufficient for most applications.
Goldman Sachs highlighted the increased adoption of Chinese AI, stressing the growing demand for these models in performing complex tasks cost-effectively. Analysts note that AI pricing based on tokens makes Chinese models more economical.
Challenges and Limitations
Despite their popularity, Chinese AI models have limitations. Anastasios Angelopoulos, CEO of Arena, observes that U.S. models are more versatile in their capabilities. Yasir Atalan from the Center for Strategic and International Studies suggests U.S. firms may seek cheaper alternatives to remain competitive.
The open-source nature of many Chinese AI models contrasts with closed systems from U.S. firms. Experts like Lian Jye Su believe this openness will boost their global adoption. Krikorian notes the growing dominance of Chinese-built open-source models.
Impact of U.S. Policies
The U.S. considers restricting Chinese AI models. Nonetheless, American tech companies have voiced support for open AI systems. Some policies, experts argue, inadvertently benefit Chinese competitors. For instance, Z.ai released its GLM-5.2 model after U.S. export controls paused Anthropic’s models.
Global Expansion Plans
Chinese AI’s success at home fuels its ambitions abroad. State support helps companies like Huawei and Tencent integrate AI into various devices. China’s President Xi Jinping advocates for open-source AI and global equity, particularly in developing nations.
As intense domestic competition drives expansion, Chinese AI startups are seeking funding for global ventures. Both countries aim to spread their AI ecosystems, protecting strategic technologies. Chelsey Tam from Morningstar notes the dual pressure on Chinese labs from both U.S. rivals and internal competition.
Despite challenges like sustaining investments, Chinese AI models remain poised to grow in the U.S. Krikorian encourages evaluations of Chinese AI for heavy workloads.
Reported by O’Brien from Providence, Rhode Island, and Kelvin Chan in Toronto for the AP.