Ant Group Turns to Homegrown Chips to Power Its AI Ambitions

In a strategic move to reduce reliance on foreign technology, Ant Group has begun training its AI models using domestically produced chips. This shift marks a significant milestone in China’s broader push for tech self-sufficiency, especially in the high-stakes field of artificial intelligence.

As a financial technology giant and affiliate of Alibaba, Ant Group has been heavily investing in AI to power its services, from intelligent credit assessments to customer service chatbots. Traditionally, like many global tech firms, it depended on foreign-made GPUs—particularly those from U.S. companies like Nvidia—for model training.

However, with rising geopolitical tensions and tightening export controls on advanced semiconductors, Ant Group is now pivoting toward Chinese-designed chips. These alternatives, while not yet as powerful as their Western counterparts, have made significant progress and are increasingly viable for large-scale AI workloads.

The benefits are twofold: cost savings and supply chain security. By using locally sourced hardware, Ant can better control costs and avoid potential disruptions caused by sanctions or export restrictions.

This move aligns with China’s national strategy to build a robust domestic semiconductor ecosystem. If successful, Ant Group’s approach could serve as a blueprint for other Chinese tech firms looking to stay competitive in the AI race without depending on foreign chipmakers.

It’s a bold bet—not just on AI, but on the future of homegrown innovation.