At the World Economic Forum (WEF) in Davos on January 21, 2026, Tang Daosheng, Senior Executive Vice President of Tencent Group and CEO of its Cloud and Smart Industries Group (CSIG), shared the company's strategic insights and industrial practices in the field of artificial intelligence. He stated, "The core of our cloud strategy is to support the operation of various different models, providing customers with tools and products that are not dependent on any specific model. We believe this is the way to return the autonomy of choosing the right model back to the client."
Tang Daosheng remarked, "When people talk about AI, they might tend to imagine it as one massive super-system, calling it AGI. But in reality, what we have are multiple different models serving different scenarios."
Research from the World Economic Forum indicates that AI adoption is accelerating in China, with 87% of Chinese enterprises planning to increase their AI investments. AI is becoming a core driver of productivity, resilience, and long-term competitiveness. However, the challenge lies in transforming innovative pilot projects into systemic change and measurable returns. Tang Daosheng pointed out that the Chinese market generally focuses on the practical benefits and cost control of AI applications. Tencent provides optimization solutions through its cloud services, stating, "We are consistently committed to optimizing the cost of using AI, making it accessible and promotable to benefit a broader population, ensuring no one is left behind."
As a globally developing internet technology company, Tencent is accelerating its AI practices and using AI systems to restructure its own operations. Tang Daosheng introduced, "Within Tencent, many programmers widely use coding tools to develop new features at a pace far exceeding the past; employees from various roles, including product managers, designers, and finance personnel, are all utilizing AI."
Currently, over 12,000 engineers within Tencent use the AI programming assistant CodeBuddy, with more than 50% of the company's new code being AI-assisted. Actual measurement data shows that CodeBuddy has reduced average coding time by over 40%. Furthermore, 90% of the code for Tencent's newly upgraded AI CLI version, CodeBuddy Code, was generated by CodeBuddy itself. This internal case has become a typical example of Tencent's push for the large-scale application of AI.
Additionally, Tencent is firmly investing in its fully self-developed full-stack Hunyuan large model, accelerating the model development process. Over the past year, it has released over 30 new models covering areas such as enhanced hybrid reasoning, image, video, and 3D generation. Among them, the newly released Hunyuan 2.0 adopts a Mixture of Experts (MoE) architecture, boasting a total parameter count of up to 406B (with 32B activated parameters), placing its reasoning capability and efficiency at an industry-leading level. Downloads for the open-source platform of the Hunyuan 3D large model have surpassed 3 million, making it recognized by global developers, creators, and the open-source community as one of the most popular open-source 3D models.
Currently, the Hunyuan large model supports efficiency improvements in over 900 business scenarios within Tencent, including Tencent Meeting, WeChat, advertising, and gaming businesses. AI is systematically transforming Tencent and generating direct, quantifiable benefits. Tencent's financial report for the third quarter of 2025 showed strong AI-driven growth: AI enabled more precise targeted advertising, improved click-through rates, and drove a 21% year-on-year increase in marketing services revenue; AI also contributed to aspects like game user engagement and matching mechanisms, leading to a 22.8% year-on-year increase in gaming revenue.
Tang Daosheng also noted that China's AI ecosystem involves numerous participants, with the number of model companies far exceeding that of other markets, and features a strong open-source atmosphere. This diversity provides customers with flexible choices, and Tencent Cloud supports clients in selecting suitable models based on different scenarios.
Using agents as the vehicle, the cloud platform as the support, and out-of-the-box tools as the means, Tencent Cloud is committed to packaging AI capabilities into solutions that can be implemented across various industries. Tang Daosheng explained that Tencent's AI has already been deployed in over 30 industries, including retail, healthcare, and education. Examples include helping retail enterprises use AI to generate 3D models and accelerate design cycles, assisting pharmaceutical companies in new drug research and development, and improving conversion rates through precision marketing. "These cases demonstrate that AI is tangibly driving industrial efficiency gains."
The systemic reshaping of various industries by AI is gradually achieving an upgrade from point-based efficiency improvements to full-chain collaboration, ultimately bringing about profound changes in cost reduction, efficiency gains, business model innovation, interaction revolution, and competitiveness restructuring. A typical case is Bambu Lab, a world-leading consumer-grade 3D printing company, which reconstructed its traditional 3D modeling workflow by utilizing Tencent's Hunyuan 3D large model. Leveraging AI's 3D generation capabilities significantly lowered the threshold of originally complex professional modeling, not only amplifying participation from ordinary users but also stimulating more personalized creative demands.
When discussing how the new generation can adapt to the educational environment shaped by AI, Tang Daosheng pointed out that the new generation learns about AI not only through schools and educational institutions but also by using readily available free AI tools like Tencent Yuanbao to ask questions and satisfy their curiosity. "I believe encouraging the new generation to maintain curiosity and skillfully use these freely available AI tools is the best way to cultivate their habit of using AI as a learning tool."
Currently, the accelerating momentum of AI development in China has become a global focus. Tang Daosheng believes that China's AI ecosystem is open, diverse, and vibrant.
This dynamism is underpinned by China's profound innovative DNA. Firstly, there is the continuous accumulation of R&D investment and talent density; in recent years, China has persistently increased investment in fundamental AI research, solidifying the foundation for AI technological innovation. Secondly, there are cutting-edge breakthroughs and architectural innovations in model technology; Chinese tech companies and research institutions closely follow global technology trends, achieving independent innovation in core areas like model architecture and training methods. Most crucially, there are extensive industrial practice scenarios; AI is accelerating its deep integration with strategic sectors like manufacturing, healthcare, finance, energy, retail, and education. The rich industrial demands and real-world business scenarios provide a globally leading fertile ground for technology to grow, allowing AI to be rapidly validated, iterated, and implemented.
Comments