Major US companies are increasingly turning to Chinese AI models to power their services, raising questions about whether China is quietly pulling ahead in the global artificial intelligence race.
Pinterest is experimenting with Chinese AI models to improve its recommendation engine, while Airbnb relies heavily on Alibaba’s Qwen for customer service. The shift comes as open-source Chinese models offer significant cost advantages over US proprietary alternatives.
The DeepSeek Breakthrough
The launch of China’s DeepSeek R-1 model in January 2025 marked what Pinterest boss Bill Ready calls a breakthrough moment. The company chose to open source the technology, sparking a wave of similar releases from Chinese competitors.
“We’ve effectively made Pinterest an AI-powered shopping assistant,” Ready told reporters. Chinese competitors now include Alibaba’s Qwen and Moonshot’s Kimi, while TikTok owner ByteDance is working on similar technology.
Pinterest Chief Technology Officer Matt Madrigal said the strength of these models is they can be freely downloaded and customized, unlike the majority of models from US rivals like OpenAI, which makes ChatGPT.
“Open source techniques that we use to train our own in-house models are 30% more accurate than the leading off-the-shelf models,” Madrigal said. The improved recommendations come at much lower cost, sometimes 90% less than using proprietary models favored by US AI developers.
Fortune 500 Adoption
Pinterest is hardly alone among US enterprises depending on Chinese AI technology. These models are gaining traction across Fortune 500 companies.
Airbnb boss Brian Chesky told Bloomberg in October his company relied “a lot” on Alibaba’s Qwen to power its AI customer service agent. He gave three simple reasons: it’s “very good”, “fast” and “cheap”.
The company uses several models, including US-based ones, hosting them securely in its own infrastructure. Data is never provided to the AI model developers, according to Airbnb.
Hugging Face Trends
Evidence of Chinese dominance appears on Hugging Face, the platform where developers download ready-made AI models from major developers like Meta and Alibaba.
Jeff Boudier, who builds products at the platform, said cost factors lead young startups to choose Chinese models over US counterparts.
“If you look at the top trending models on Hugging Face, typically Chinese models from Chinese labs occupy many of the top 10 spots,” he said. “There are weeks where four out of five top training models on Hugging Face are from Chinese labs.”
In September, Qwen topped Meta’s Llama to become the most downloaded family of large language models on Hugging Face.
Meta released its open-source Llama AI models in 2023. Until DeepSeek and Alibaba’s releases, they were considered the go-to choice for developers working on custom applications. But the release of Llama 4 last year left developers underwhelmed.
Meta has reportedly been using open-source models from Alibaba, Google, and OpenAI to train a new model set for release this spring.
Chinese Models Catch Up
A report published last month by Stanford University found Chinese AI models “seem to have caught up or even pulled ahead” of global counterparts in both capabilities and user adoption.
“That’s not the story anymore,” Boudier said. “Now, the best model is an open-source model.”
Former UK deputy prime minister Sir Nick Clegg told the BBC he felt US firms were overly focused on pursuing AI that may one day surpass human intelligence. Last year, Sir Nick left his post as head of global affairs at Meta. Boss Mark Zuckerberg has committed billions of dollars to achieving what he calls “superintelligence”.
Some experts now call these ambitions vague and ill-defined, giving China an opening to dominate the open-source AI space.
“Here’s the irony,” Sir Nick said. In the battle between “the world’s great autocracy” and “the world’s greatest democracy”, China and America, China is “doing more to democratize the technology they’re competing over”.
Government Support vs Revenue Pressure
The Stanford report suggested China’s success in developing open-source models could be partly explained by government support.
US companies like OpenAI face intense pressure to increase revenue and become profitable. The company is now turning to ads to help achieve that goal.
OpenAI released two open-source models last summer, its first in years. But it has poured most resources into proprietary models to generate revenue.
OpenAI boss Sam Altman said in October the company has invested aggressively in securing computing power and infrastructure deals.
“Revenue will grow super fast, but you should expect us to invest a ton in training, in the next model and the next and the next and the next,” he said.
What It Means
The shift to Chinese AI models represents a significant development in the global technology landscape. US companies are choosing performance and cost savings over domestic alternatives, even as geopolitical tensions persist.
Whether this trend continues depends on how US companies respond to the open-source challenge and whether they can match the cost-effectiveness of Chinese competitors while maintaining their technological edge.
