China Surges Ahead In AI Race With DeepSeek While U.S. Struggles To Keep Pace
Microsoft warns that Chinese AI companies are rapidly gaining ground globally, using low-cost models and heavy state support to reach markets beyond the West.
Brad Smith, Microsoft president, highlighted DeepSeek’s rapid adoption in emerging markets, noting that China’s open-source models are now competitive and often cheaper than U.S. offerings.
“They benefit from subsidization by the Chinese government. They benefit from subsidies that enable [them] to basically undercut American companies based on price.”
DeepSeek Captures Markets Where U.S. Tech Is Limited
Data from Microsoft shows that DeepSeek’s R1 model, launched last year, has quickly expanded in countries where American firms have limited access.
The company now holds 56% of the AI market in Belarus, 49% in Cuba, and 43% in Russia.
In Africa, it commands 18% of Ethiopia and 17% of Zimbabwe, driven by affordability and open access.
Source: Microsoft
The model’s open-source approach, including free usage and adaptability, has accelerated AI adoption in regions often overlooked by Western companies.
Smith said this growth reflects a broader shift in the global AI landscape.
Is Open Source Outpacing Locked Ecosystems
While U.S. companies like OpenAI, Google, and Anthropic maintain paid, controlled ecosystems, Chinese platforms such as DeepSeek and Meta’s Llama benefit from open-source accessibility.
Smith warned that without significant investment in infrastructure, data centres, and energy support, this divide could deepen global inequality.
“If we rely on private capital flows alone, I don’t think that will be sufficient to compete with a competitor that is subsidised to the degree that Chinese companies often are.”
He called for international development banks and governments to provide funding to support AI adoption in poorer regions.
Bright Simons, an AI analyst from Ghana’s IMANI think-tank, said Chinese AI tools are now the go-to for many Africans.
He explained,
“Africans can’t afford very expensive solutions apart from open source, so you have to go to [Meta’s] Llama or Chinese options.”
He also noted local efforts such as Masakhane and InkubaLM, but Chinese models remain dominant in accessibility and reach.
Global AI Use Shows Wide Divide Between North And South
Microsoft’s AI Diffusion Report 2025 shows that 24.7% of people in the global north used AI by the end of 2025, compared with just 14.1% in the global south, leaving a worldwide average of 16.3%.
Source: Microsoft
Smith called the gap a “cause for concern,” warning that without targeted investment, inequality will widen further.
“American companies, as well as Western governments, cannot ignore Africa. If they do, they are closing their eyes to the future of the world more broadly, and I think that would be a grave mistake.”
Strong Models Don’t Guarantee Widespread Adoption
Even in technologically advanced countries, public use of AI does not always match technical capability.
By the end of 2025, about 28% of working-age Americans had used generative AI, placing the U.S. 24th globally.
Meanwhile, South Korea climbed from 25th to 18th, boosted by clearer AI policies, improved language support, and public engagement through cultural trends such as Ghibli-style graphics on social media.
Source: Microsoft
China’s Strategy Focuses On Accessibility And Low Costs
DeepSeek has leveraged its MIT-licensed model and free chatbot to lower both financial and technical barriers.
Adoption in North America and Europe remains limited, but the model has spread quickly across China, Russia, Iran, Cuba, Belarus, and parts of Africa, where usage rates can be two to four times higher than in many Western countries.
Smith emphasised that the U.S. retains advantages in trust and advanced chip access but warned that price competition and global reach are increasingly critical.
DeepSeek’s next major model is expected to launch just before Lunar New Year, signalling continued momentum in regions where American firms have struggled to make inroads.