The U.S. restricted China's access to AI chips. That didn't stop DeepSeek

China pushes the boundaries of artificial intelligence advancement through open-source innovation and resource efficiency, bypassing some U.S. sanctions

We may earn a commission from links on this page.
Image for article titled The U.S. restricted China's access to AI chips. That didn't stop DeepSeek
Photo: Artur Widak/NurPhoto (Getty Images)

This story incorporates reporting from China Briefing, The Washington Post and The Associated Press on MSN.com.

China’s AI sector is advancing rapidly, led by innovation and resource-efficient practices. Pioneering companies like DeepSeek have significantly contributed to these advancements with open-source models that challenge conventional AI development methods. These models have reduced reliance on traditional chip-intensive processes, allowing China to maintain its competitive edge despite U.S. chip restrictions. This progress marks a shift in China’s AI industry toward sustainability and collaboration, driving it to the forefront of international discussions on AI governance and policy.

DeepSeek, a notable player in China’s AI landscape, exemplifies this shift with its cutting-edge R1 AI model. The company’s approach has captivated the tech community by cutting costs and minimizing dependency on high-performance chips. By leveraging open-source strategies, DeepSeek and other companies offer free access to their models worldwide, fostering a collaborative ecosystem that supports innovation beyond national borders.

Advertisement

One of DeepSeek’s significant contributions is its R1-Zero model, which challenges the conventional belief in the necessity of supervised fine-tuning during AI model development. By eliminating this stage, DeepSeek has streamlined the AI training process — showing that large models can be trained efficiently without extensive computational resources.

Advertisement

The growing trend of open-source AI models in China, from startups like Minimax to tech giants such as Alibaba, signals a strategic movement toward shared development and cost efficiency. This trend allows Chinese companies to make impressive strides without the high-performance chips that are restricted by U.S. sanctions. By emphasizing collaboration and open access, China not only bypasses some effects of these restrictions but also actively participates in reshaping global AI research dynamics.

Advertisement

China’s advancements in AI, driven by resource efficiency and open-source contributions, influence the competitive dynamics between nations. These developments are spurring discussions around the need for a revised global AI policy, as China’s approach disrupts traditional models of AI creation and governance. As these conversations continue, China’s innovations stand as a testament to the potential of alternative strategies in overcoming geopolitical and technological barriers.

Looking ahead, the focus will likely remain on how well China can sustain this momentum and how their approach might influence international policy frameworks. As of January 2025, the AI landscape continues to evolve rapidly, with China prominently positioned to shape its trajectory.

Advertisement

Quartz Intelligence Newsroom uses generative artificial intelligence to report on business trends. This is the first phase of an experimental new version of reporting. While we strive for accuracy and timeliness, due to the experimental nature of this technology we cannot guarantee that we’ll always be successful in that regard. If you see errors in this article, please let us know at qi@qz.com.