Summary of Key Points
After Wu Yongming took over as CEO of Alibaba in 2023, the company shifted from having almost no AI investments (only 3 between 2019 and 2021) to making a bold strategic move: the number of investments surged within two years (11 in 2023 and 34 in 2025). The approach changed from waiting for data to be generated before investing in the mid-to-late stages to making early investments as soon as companies were established. The three main entities within Alibaba Group—Ant Group, Alibaba itself, and Alibaba Cloud—have clear roles, covering all aspects of the large model ecosystem, including the six leading AI companies (referred to as the “six dragons”) and the entire value chain. They have created a closed-loop strategy that involves “investing in models → purchasing computing power → generating revenue.” However, this approach also comes with risks such as potential redundancy in investments, high valuations of startups, and uncertainty about strategic continuity. Essentially, it’s a defensive “insurance strategy” to ensure Alibaba doesn’t miss out on the opportunities of the AI era.
Detailed Analysis
1. The Rapid Shift from Inactivity to Aggressive Investment
Alibaba was relatively inactive in the AI sector before: from 2019 to 2021, it only invested in three AI companies, which is negligible compared to the global AI boom sparked by ChatGPT. On the third day of Wu Yongming’s tenure, he warned that failing to keep up with AI could lead to being replaced by new technologies. Following this warning, Alibaba accelerated its investment pace significantly, investing in 34 companies in 2025, more than in the previous ten years combined. More importantly, the company’s investment strategy has changed dramatically: it now invests at the earliest stages, even before companies have generated user data or proven profitability (angel and Pre-A rounds are now common). This change reflects a clear sense of urgency to avoid missing out on potential opportunities.
2. Clear Division of Labor among the Three Entities
The three main entities within Alibaba Group work together efficiently:
- Ant Group: The most active “scout team,” with 48 investments, mostly in the early stages (7-8 each in angel, Pre-A, and A+ rounds), searching for promising startups across various AI fields.
- Alibaba itself: The “main force,” focusing on mid-to-late-stage investments, particularly in large models and infrastructure (9 strategic investments and 6 in Series C rounds), such as Zhipu and Yuezhi Dianmian.
- Alibaba Cloud: Acting as a bridge, making investments around its cloud services (for example, in companies that use Alibaba Cloud’s AI solutions).
Some companies have received investments from all three entities (e.g., Zhipu and Jiushi Intelligence), ensuring comprehensive coverage.
3. Broad Coverage across Various AI Fields
Alibaba has made a bold bet on multiple AI sectors:
- Large Models: It has invested in all six leading large model companies, including those not even invested in by Tencent or Baidu, and has made additional rounds of investment (rare in the venture capital community). Its scope has expanded from computer vision (e.g., Megvii and SenseTime) to AI-generated content (AIGC), embodied intelligence (8 companies), and cloud computing (9 companies).
- Chip Technology: Investments include ChangXin Memory (a leading DRAM manufacturer), Suoyuan Technology (comparable to NVIDIA’s A100), and Cambricon, covering storage, training, and inference technologies.
- Embodied Intelligence: It has invested in companies developing quadrupedal and humanoid robots, as well as those focused on bipedal control technology.
Over 45% of the invested companies received funding within one year of establishment; for example, Yuezhi Dianmian received an A-round investment from Alibaba in April 2024.
4. The Computing Power Closed Loop: Turning Investments into Revenue
Alibaba’s business model relies on a closed-loop approach where investments lead to new customers and generate revenue. For instance, Minimax purchased $58.4 million in computing power from Alibaba in the first three quarters of 2025, which means the money invested was recouped through sales of computing resources. This isn’t just a financial investment; it’s also about acquiring customers for Alibaba Cloud, thus doubling the benefits.
5. The Challenges of this Strategy
While this strategy aims to maximize returns, it also brings risks:
- Redundancy: Some of the 16 AIGC companies (such as Zhipu and BaiChuan) are competing with open-source large models, which could lead to integration or exit issues in the future.
- High Valuations: The high valuation of some startups (e.g., Yuezhi Dianmian, valued at $20 billion with Alibaba holding 36% of the shares) depends on the success of the closed-loop strategy.
- Dependence on Cloud Services: Using cloud credits instead of cash may reduce the independence of invested companies.
- Strategic Uncertainty: With key executives leaving in 2024, there’s concern about potential changes in future investment directions.
In essence, Alibaba’s approach is a defensive one, aiming to secure a position in the AI industry by investing in various technologies without knowing which will succeed. By diversifying its investments and collaborating among its entities, it hopes to remain competitive during the industry’s transformation.
Conclusion
Alibaba’s aggressive AI investment strategy is a desperate attempt to catch up with the pace of innovation. While it aims to gain a foothold in the AI ecosystem through broad coverage and collaboration, it also faces challenges such as potential redundancy and high valuations. The ultimate success of these investments will depend on how well they contribute to Alibaba’s competitiveness after the industry consolidates. For outsiders, this reflects the anxiety of tech giants: if they fail to keep up, they risk being left behind in the rapidly evolving technology landscape.