RECENT ABSTRACTS

Summary of

Examining China’s AI Financing (Rhodium Group, Sept. 2026) [rhg.com]

This report examines how Chinese AI companies are funding their rapid expansion and whether that growth is financially sustainable compared with the United States. [rhg.com]

Main conclusions

1. China's AI investment boom is growing rapidly, but remains much smaller than the US.
Rhodium Group estimates that China's AI infrastructure spending will reach about 932 billion yuan ($139 billion) in 2026, more than doubling from the previous year, and could exceed 1.2 trillion yuan ($193 billion) in 2027. However, that is still only around 15-20% of US AI investment levels. [rhg.com]

2. Financing, not just chips, is becoming a major constraint.
Much attention has focused on access to advanced semiconductors, but the report argues that financing is also a critical bottleneck. Chinese AI firms are spending heavily on data centers and computing infrastructure while generating insufficient cash flow from their businesses. [rhg.com]

3. Chinese AI firms face the same basic problem as US firms: huge capex, weak cash generation.
Large Chinese technology firms are investing aggressively in AI while free cash flow has deteriorated. The combined free cash flow of Alibaba, Tencent, and Baidu turned negative in the first half of 2026 as investment spending surged. [rhg.com]

How China finances AI

The report identifies four major groups:

• Hyperscalers: Alibaba, Tencent, Baidu
• Telecom and infrastructure firms: Huawei, China Mobile, China Telecom, China Unicom
• Frontier AI labs: Zhipu AI, MiniMax
• Independent data center operators [rhg.com]

Their financing sources differ significantly:

Hyperscalers

• Rely increasingly on equity financing and bank borrowing.
• Have reduced share buybacks and raised new capital to fund AI expansion.
• Example: Alibaba announced a large share placement dedicated to AI investment. [rhg.com]

Telecom firms and Huawei

• Mostly fund AI spending through operating cash flow generated by their existing businesses.
• Carry relatively modest levels of new debt. [rhg.com]

Frontier AI labs

• Generate little or no positive operating cash flow.
• Depend heavily on venture capital, private equity, IPOs, and share issuance.
• Their ability to expand depends strongly on investor enthusiasm and stock market conditions. [rhg.com]

Data center operators

• Use the broadest mix of financing, including:
• loans,
• bonds,
• equity,
• REITs,
• asset-backed securities,
• leasing arrangements. [rhg.com]

Key difference from the United States

The report highlights a major contrast:

• US AI firms increasingly rely on bond markets and private credit.
• Chinese AI firms rely much more on equity financing and bank loans.
• China's corporate bond market has not yet become a major source of AI funding. [rhg.com]

What could slow China's AI expansion?

The authors argue that China's AI buildout depends on:

1. Continued access to financing through equity markets.
2. Ongoing profitability from large tech companies' non-AI businesses.
3. Government support and industrial policy.
4. Availability of advanced chips and computing infrastructure. [rhg.com]

They also note that Chinese government support is likely to remain more focused on semiconductor development than on directly funding frontier AI labs. [rhg.com]

Bottom line

The report's central message is that China is investing enormously in AI and accelerating quickly, but its AI ecosystem remains substantially smaller than the US and faces significant financing challenges. Chinese AI companies currently depend heavily on equity investors and bank lending because revenues and profits are not yet sufficient to fund their ambitious expansion plans. Whether China's AI boom continues at its current pace will depend heavily on capital market conditions and state support. [rhg.co