China’s AI venture playbook is going global

Global attention might be on the US at present thanks to the football world cup, but there appears to be a more consequential contest unfolding elsewhere in artificial intelligence.

China’s largest technology groups are adapting the corporate VC strategies that helped them dominate the mobile internet era in the country to shape the emerging AI ecosystem more globally. Rather than treating venture investing just as a source of financial returns, companies, such as Tencent and Alibaba, are also using it to secure access to critical technologies, talent and international markets as AI competition intensifies.

Tencent’s reported efforts to purchase Singapore-based autonomous AI developer Manus illustrate this shift. According to reports, the company is leading discussions with Manus’ original investors, including HSG and ZhenFund, after Meta’s proposed $2bn acquisition was derailed by Chinese regulators.

The reported intervention reflects Beijing’s growing determination to retain influence over strategically important AI assets, particularly where advanced technologies risk falling under foreign ownership or becoming subject to export-control considerations.

The move also sits within a broader investment strategy.

Tencent is already an investor in DeepSeek, the Chinese large language model developer that has rapidly emerged as one of the country’s leading AI companies. Bloomberg reported that DeepSeek, having raised about $7bn at a roughly $50bn valuation in its first external funding round this year, is now seeking a further $1.5bn at an estimated $71bn valuation ahead of a potential flotation. The company’s cloud services run on Huawei chips, highlighting China’s efforts to build an increasingly self-sufficient AI technology stack despite US semiconductor restrictions.

If your company runs a venture building programme, please take a few minutes to tell us about it through this quick survey.

Alibaba has adopted a similar approach. The ecommerce group recently backed Singapore-based AI video platform PixVerse in a $439m series C financing that valued the company at more than $2bn. Like Tencent’s investments, the transaction reflects a focus on backing AI application companies with international growth prospects rather than limiting investment to domestic markets.

This approach echoes the strategy that transformed China’s internet sector a decade ago. Tencent and Alibaba themselves benefited from early strategic investors before becoming prolific backers of startups, using minority investments to build ecosystems around their platforms. Today, they are applying the same model to AI, where autonomous agents, robotics and physical AI could define the next computing platforms.

Singapore has become an increasingly important base for this expansion. Its position as an international financial centre allows Chinese-founded AI companies, such as Horizon Robotics’ agentics spin off Acrab, to access global capital while maintaining links to China’s technology ecosystem, making it an attractive location for cross-border investment and corporate structuring.

The implications extend beyond individual transactions. The US remains the leader for now in frontier AI research, but China is demonstrating that venture investing can be a strategic tool as much as a financial one. After all, the US lost China’s domestic market in the 2010s through a combination of Chinese state support and local competition driving improvements through CVC investing. Now it risks losing other parts of the world if it is also pushed out of an increasing number of Singapore-based deals.

If the mobile internet era showed the power of combining corporate capital with ecosystem building, the AI era may do the same on a global scale. The next phase of competition will be shaped not only by who develops the most capable models, but by who builds the strongest network of companies around them.

Be a part of the conversation if you’re a CVC in the region looking to scale investment and innovation in the local ecosystem