US Companies Adopt Kimi K3 AI Models Despite Growing Government Backlash 2026

US companies are increasingly adopting Chinese AI models like Kimi K3, and the US government is not happy about it. The Trump administration appears ready to crack down on these open-weight models, but banning them could backfire badly.

Why US Companies Are Turning to Chinese AI Models

The appeal of Chinese AI models like DeepSeek V4 and Kimi K3 is simple: they are open-weight, meaning companies can download and run them on their own servers. This drastically reduces inference costs and keeps all data private.

Brian Armstrong, CEO of Coinbase, has publicly stated that his company uses models like GLM-5.2 and Kimi K2.7 in production locally. This allows them to maintain control over sensitive data while benefiting from advanced AI capabilities.

The US Government's Pushback

The US Department of Commerce has already considered adding several Chinese AI startups, including DeepSeek, to its Entity List. This would limit their access to US-developed hardware and sensitive technology.

More recently, there have been discussions about drafting an executive order that would hold US companies accountable for security breaches resulting from using Chinese models in their systems. The goal is to discourage the use of these models as much as possible. The background to this story is set out in Kimi K3: China's AI Model Tops Frontend Code Arena.

Why Banning Chinese AI Models Is a Terrible Idea

Banning Chinese AI models would be a terrible idea for several reasons. First, it would deprive US companies of cost-effective and privacy-preserving alternatives to domestic models. Second, it would stifle innovation by limiting access to cutting-edge AI technology. The longer history behind the story is traced in Elon Musk Calls Kimi K3 'Impressive' as Open-Source AI Race Heats. More on the surrounding debate is collected in Kimi K3 Was Just the Warning: Qwen 3.8 Confirms China Is Closing the AI Model.

Third, it would be difficult to enforce, as open-weight models can be downloaded and run locally, making them hard to regulate. Finally, it could lead to retaliation from China, further escalating tensions. The background to this story is set out in Beam: US Open-Weight AI Strategy to Counter Chinese Models. A related thread runs through Why Britain Should Reject Cheap Chinese AI Models Now. Earlier coverage in Open-Weight AI Models Face US Scrutiny as China Pushes Control shows how the argument developed.

• Chinese AI models offer significant cost savings and data privacy benefits.
• US companies like Coinbase are already using them in production.
• The US government is considering measures to restrict their use.
• Banning these models could harm US competitiveness and innovation.
• Enforcement would be challenging due to the open-weight nature of the models.

Comparing Open-Weight and Proprietary AI Models

What This Means for US Businesses

US businesses need to stay informed about potential regulations that could affect their use of Chinese AI models. While these models offer clear benefits, the regulatory landscape is uncertain. A closer look at the same subject appears in Chinese AI Models Qwen3.8-Flash and DeepSeek V4-Flash Disrupt Pricing. Anyone following the wider picture can start with Chinese AI Models Show Alarming Misalignment in New Tests. The same theme surfaces in Chinese AI Firms Now Race Toward Colossal Models After Small-Model Lead, which takes a different angle.

Companies should consider diversifying their AI stack and preparing for possible restrictions. They should also engage in advocacy efforts to ensure that any regulations are balanced and do not unnecessarily hinder innovation. Context on how this played out before sits in German Search Engine Ecosia Ditches Mistral for Chinese Open-Source AI Models. A companion piece, UK Economy Grows 0.1% in May Despite Iran War Impact, approaches the same subject from another angle.

FAQ