US Threatens Sanctions on Chinese AI Models Over IP Theft Claims
Treasury Secretary Scott Bessent warned Washington could sanction Chinese open-source AI models over alleged intellectual property theft, raising the stakes in an already tense technology rivalry.
US Treasury Secretary Scott Bessent said on Tuesday that Washington will scrutinize Chinese open-source AI models for signs of intellectual property theft. Speaking on Fox Business, Bessent said the administration supports open-source development but will not accept foreign companies misappropriating proprietary technology from American AI firms. He confirmed that the US has legal authority to sanction Chinese entities if evidence shows their models were built on stolen software or research.
The statements follow reports that the Trump administration is weighing options to limit the influence of foreign open-weight systems, a significant policy shift that moves beyond hardware export controls to target software architectures directly. Recent releases from Chinese firms, including Moonshot AI's Kimi K3, have closed meaningful performance gaps with top US systems such as Anthropic's Claude Opus and OpenAI's flagship models. American labs have raised concerns that competitors are using model distillation to replicate their capabilities at a fraction of the cost, undermining the financial case for continued frontier research investment.
Model distillation lets developers train smaller, more efficient systems using outputs from larger models. Some American labs treat this as unauthorized copying when done without permission. Others push back. Microsoft CEO Satya Nadella noted the contradiction in large labs claiming fair use rights to train on public internet data while seeking to restrict downstream distillation. Hugging Face CEO Clem Delangue argued that distillation is standard practice across the tech sector, including within the US, and that China's technical momentum reflects strong local research talent and collaborative development culture, not imitation.
The intellectual property debate is not confined to US-China tensions. American developers face their own copyright disputes over training data, illustrated by Anthropic's recent approval to begin paying authors under a $1.5 billion settlement involving downloaded books. Enforcing sanctions against foreign open-weight models also presents a practical challenge: once published, open-source code spreads rapidly across global repositories, peer-to-peer networks, and decentralized cloud infrastructure, making containment difficult.
For Africa's technology sector, the conflict carries real operational risk. Founders and engineers across Lagos, Nairobi, Johannesburg, and Accra depend heavily on open-source and open-weight AI models to build localized tools. Persistent foreign exchange volatility, high inflation, and steep US API subscription costs make open models one of the few cost-effective paths to AI deployment without draining scarce foreign currency. If sanctions restrict access to Chinese open-source architectures, African engineering teams could find their options shrinking fast. African governments currently drafting AI policy, including Nigeria's national AI strategy and Kenya's digital governance framework, will also face pressure to decide how to handle international code restrictions without compromising technical sovereignty or closing off access to critical infrastructure.
When superpowers turn software into a weapon, it is developers in emerging markets who absorb the shrapnel.
Source: TechCrunch
Written by
StacyAI-assisted news curation. Every story is reviewed by our editors before publication.
