Pinterest is experimenting with Chinese AI models to power its recommendation engine, highlighting a growing trend among US tech companies to adopt foreign artificial intelligence technology.
“We’ve effectively made Pinterest an AI-powered shopping assistant,” said Bill Ready, the company’s CEO. The San Francisco-based platform uses AI to suggest everything from wacky DIY projects to new fashion trends for its hundreds of millions of users worldwide. While Pinterest could rely on American AI labs, the launch of China’s DeepSeek R-1 model in January 2025 has introduced new options for tech firms looking for performance and affordability.
Matt Madrigal, Pinterest’s Chief Technology Officer, said the open-source nature of Chinese models allows the company to train its own in-house systems at lower cost. “Open source techniques that we use to train our own in-house models are 30% more accurate than the leading off-the-shelf models,” he said. Madrigal added that using these models can reduce expenses by up to 90% compared with proprietary US alternatives.
Chinese AI models have gained traction beyond Pinterest. Airbnb, for example, relies heavily on Alibaba’s Qwen model to power its AI customer service agent. CEO Brian Chesky told Bloomberg the choice was driven by three factors: “very good, fast, and cheap.”
The popularity of Chinese models is evident on Hugging Face, a platform used by developers to download ready-made AI models. Jeff Boudier, who oversees product development there, said Chinese models frequently dominate the top trending charts. “There are weeks where four out of five top training models on Hugging Face are from Chinese labs,” he said. In September 2025, Alibaba’s Qwen surpassed Meta’s Llama to become the most downloaded family of large language models on the platform.
US companies are cautious about security and data privacy when using these models. Airbnb, for example, hosts all AI models in its own infrastructure, ensuring that user data is not shared with model developers.
Experts say China’s success with open-source AI is partly due to government support and the ability to focus on democratizing the technology. Former UK deputy prime minister Sir Nick Clegg, who recently left Meta, argued that while US firms chase “superintelligence,” China is advancing open-source models accessible to a wider range of developers.
A recent Stanford University report found that Chinese AI models have caught up with or even surpassed global competitors in terms of capability and adoption. In contrast, US firms like OpenAI are increasingly focused on profitability, directing resources to proprietary models and infrastructure deals to secure revenue.
The rise of Chinese AI in US companies reflects both cost pressures and the practical advantages of open-source solutions. For Pinterest and other firms, these models offer faster development, improved accuracy, and more flexible deployment, signaling a shift in how American tech firms approach AI development in an increasingly competitive global landscape.
