In Washington, discussions about imposing restrictions on the use of Chinese open-source artificial intelligence models have intensified once again. The catalyst for this process was the release of the powerful Kimi K3 model by Moonshot AI, which, in my assessment, demonstrated that Chinese developers are capable of creating solutions comparable to the best Western counterparts.

Scenarios and Tools of Pressure

As of now, none of the proposed measures have achieved official policy status. However, based on an analysis of the current dynamics, the administration is considering a whole range of measures: from adding Chinese AI companies to export blacklists—which would require special permission from the U.S. Department of Commerce for any interaction—to issuing public warnings about potential national security threats posed by these models.

Earlier, more stringent options, including an executive order placing responsibility on American companies for the safe use of open-source models, were rejected due to the risk of slowing down the industry's development pace. Currently, the focus is on less radical but equally effective tools: restrictions in government procurement and targeted warnings from regulators.

Expert Opinion and White House Reaction

Dean Ball, former senior advisor on AI policy at the White House, now heading the Strategic Futures division at OpenAI, expressed the view that the best strategy for the Trump administration would be to create "regulatory risk" for the use of Chinese open-source models. According to him, this is not a recommendation but a forecast based on the logic of competitive struggle.

However, the President's advisor on AI, David Sacks, categorically rejected this approach, calling "weaponizing regulatory uncertainty" unacceptable. He noted that leading developers of closed systems, such as OpenAI and Anthropic, are effectively forming a duopoly and may be interested in limiting competition from open-source solutions. This internal debate within the White House underscores just how complex and multifaceted the issue of AI regulation is.

China Does Not Stand Idle

In parallel with American initiatives, Beijing is also considering introducing its own export restrictions on advanced AI technologies. China's Ministry of Commerce has already held consultations with Alibaba, ByteDance, and Zhipu regarding control over the transfer of training data and model weights abroad. This is a mirror response that only exacerbates the global technological divide.

The situation is further complicated by mutual accusations. Anthropic has repeatedly claimed massive campaigns to "distill" its Claude model by Chinese startups, including DeepSeek, Moonshot, and MiniMax. Alibaba, in turn, has banned its employees from using Anthropic products, citing security risks.

My analysis: We are witnessing not just a trade war, but a fundamental struggle for control over the future of AI. Neither side is yet ready for radical steps, but creating "regulatory landmines" is becoming a key tool. The release of Kimi K3 only accelerated this process, showing that the technology gap is narrowing, meaning pressure from the U.S. will only increase, regardless of who sits in the Oval Office.