A new investigation by the Associated Press reveals that Chinese artificial intelligence models are making significant inroads into the United States market, challenging the dominance of domestic tech giants like OpenAI and Anthropic. The report documents real-world adoption by prominent US companies, driven by a combination of competitive performance and significantly lower costs.
The findings highlight a shifting dynamic in the global AI race. While the US has traditionally led in the development of frontier models, Chinese companies are rapidly closing the capability gap and aggressively competing on price. This trend is forcing US businesses to re-evaluate their AI strategies and raising concerns in Washington about the long-term implications for American technological leadership.
Real-World Adoption by US Tech Leaders
The AP investigation provides concrete examples of US companies integrating Chinese AI models into their operations. One notable case is Mozilla, the organization behind the Firefox web browser. According to the report, Mozilla’s Chief Technology Officer switched to using Moonshot AI’s Kimi K3 model within days of its launch, citing its impressive performance and open-weight availability.
Another significant example is Coinbase, the largest cryptocurrency exchange in the US. The AP reports that Coinbase is in the process of switching some of its AI workloads to Chinese models in an effort to trim costs. This trend has been building for months: as we reported in July, Chinese AI models have been capturing the US developer market with cost as the primary driver. This move by a major financial technology company underscores the growing appeal of Chinese AI solutions for cost-conscious enterprises.
The adoption of Chinese models is not limited to these high-profile examples. The AP investigation suggests a broader trend of US developers and startups experimenting with and deploying models from companies like DeepSeek, Z.ai (Zhipu AI), and Moonshot AI. These models, particularly Z.ai’s GLM-5.2 and Moonshot’s Kimi K3, are increasingly described by industry experts as being “nearly as intelligent as frontier models from OpenAI and Anthropic.”
The Cost Advantage
The primary driver behind the adoption of Chinese AI models in the US is cost. Chinese companies are offering their models at a fraction of the price charged by their American counterparts. This aggressive pricing strategy is making advanced AI capabilities accessible to a wider range of businesses, from startups to large enterprises looking to optimize their operations.
The cost advantage is particularly significant for applications that require processing massive amounts of data or generating large volumes of text. For these use cases, the API costs associated with models like GPT-4 or Claude 3 can quickly become prohibitive. Chinese models offer a compelling alternative, providing comparable performance at a significantly lower price point.
Furthermore, the availability of open-weight models like Kimi K3 allows companies to self-host the AI, eliminating API costs entirely and providing greater control over data privacy and security. This self-hosting capability is a major selling point for enterprises with strict compliance requirements or those looking to build custom AI solutions.
Distillation Accusations and Geopolitical Tensions
The rapid rise of Chinese AI models has not gone unnoticed in Washington, and the AP report highlights the growing geopolitical tensions surrounding the technology. The Trump administration has recently accused Moonshot AI of engaging in “covert but not necessarily illegal” distillation of Anthropic’s Fable model.
Distillation is a process where a smaller, cheaper model is trained using the outputs of a larger, more capable model. While this practice is common in the AI industry, the accusation suggests that Chinese companies may be leveraging American innovations to accelerate their own development. These allegations have fueled calls for stricter export controls and regulations to protect US intellectual property.
The AP report also notes that Treasury Secretary Scott Bessent has warned that sanctions against Chinese AI companies are “on the table.” The State Department previously sent a global diplomatic cable warning allies about Chinese AI distillation campaigns by DeepSeek, Moonshot AI, and MiniMax. This threat of regulatory action adds a layer of uncertainty for US companies considering adopting Chinese models. The potential for sudden disruptions in access or legal complications could deter some businesses from fully committing to Chinese AI solutions.
The Future of the US AI Market
The AP investigation paints a picture of a rapidly evolving US AI market, where Chinese models are emerging as serious competitors to established domestic players. The combination of competitive performance, aggressive pricing, and open-weight availability is proving highly attractive to US businesses.
However, the geopolitical context cannot be ignored. The ongoing tech war between the US and China, coupled with accusations of intellectual property theft and the threat of sanctions, creates a complex and volatile environment for AI adoption. US companies must carefully weigh the economic benefits of using Chinese models against the potential regulatory and reputational risks.
As the AI race continues to accelerate, the competition between US and Chinese models will only intensify. The ability of Chinese companies to maintain their cost advantage while continuing to improve model capabilities will be crucial to their long-term success in the US market. Meanwhile, US tech giants will need to innovate rapidly and potentially adjust their pricing strategies to defend their market share against this new wave of competition.