Decoding China's AI Ambitions: WAIC 2025

The World Artificial Intelligence Conference (WAIC) in Shanghai is rapidly evolving beyond a mere tech showcase. It’s becoming a strategic platform for China’s industrial policy and a barometer of global technological competition.

AI as a Cornerstone of the New Industrial Era

The 2025 WAIC theme, “Intelligent Empowerment, Generative Future,” signals a pivotal shift. The focus is now firmly on integrating AI into the real economy, driving a new wave of industrialization, which aligns perfectly with China’s “new productive forces” strategy. This strategy aims to move away from traditional growth models towards innovation-driven, high-value industries.

China sees WAIC as a key platform for building a globally competitive and self-reliant AI ecosystem. This dual objective positions the conference as both a mobilizer, aligning domestic resources, and a projector, showcasing China’s AI-driven progress on the world stage.

To illustrate this evolution, consider the changing themes of WAIC:

• 2023: “Intelligent Connectivity, Generative Future” – Focused on large language models (LLMs), AIGC, and the metaverse.
• 2024: “Global Collaboration for Shared Future” – Emphasized multimodal models, embodied AI, and data.
• 2025: “Intelligent Empowerment, Generative Future” – Prioritized industrial applications, digital twins, and AI for Science.

This progression clearly demonstrates a move from pursuing fundamental model capabilities to emphasizing quantifiable industrial output and economic value. The shift in terminology from abstract concepts to tangible applications underscores WAIC’s role in reflecting and driving China’s strategic economic priorities.

Translating Policy into Action

WAIC has transcended its role as a tech exhibition, becoming a policy instrument. The alignment of the conference theme with high-level economic strategies highlights a deliberate, top-down design. The aim is no longer simply to showcase technological achievements but to mobilize the AI ecosystem towards the intelligent upgrading of China’s industrial base.

WAIC 2025 marks a strategic shift in China’s AI strategy. It’s a move from “catching up” in basic technologies to “leading” in industrial applications. This is a direct response to geopolitical pressures, particularly US-led restrictions on advanced semiconductor exports. China is strategically reshaping the competitive landscape by focusing on its unique advantage: a vast and comprehensive industrial base.

China is positioning itself as a leader by leveraging its massive domestic market and manufacturing prowess to build a globally relevant AI leadership position, bypassing direct confrontation in semiconductor hardware.

Cutting-Edge Technologies in Focus

WAIC 2025 showcases a pragmatic approach to technology, driven by real-world problem-solving, supply chain resilience, and commercial value creation.

The Evolution of Foundation Models

The “parameter race” has subsided, replaced by a more refined approach centered on efficiency, multimodal capabilities, and vertical applications. The consensus is shifting towards prioritizing return on investment (ROI) over sheer model size.

The “Pangu-Σ” series of models exemplifies this trend. These models emphasize multimodal fusion and specific application scenarios, such as high-precision defect detection in industrial quality control. This signals a move towards specialized models designed for cost-effective deployment on edge computing devices or enterprise environments while addressing the shortcomings of generalized large models in commercial applications. These models try to solve problems of cost and scalability as there is an increasing concern about the economic viability of foundation models. Many companies are trying to leverage transfer learning from large models to fine tuned smaller models, which is a cost effective path to building value.

Embodied Intelligence and Industrial Robots

Embodied intelligence, particularly humanoid robots, is emerging as a core pillar of the “intelligent empowerment, generative future” strategy. The focus is shifting from entertainment features to real-world operational capabilities in manufacturing and logistics.

Exhibitors like “RoboForge” are showcasing humanoid robots in industrial settings, showing a 30% efficiency increase in specific tasks. This highlights the transition from technological feasibility to economic viability. This progress is driven by the integration of advanced robotics hardware (such as high-precision joints) with AI models, enabling robots to perform intricate tasks in complex environments. It is becoming clear that robotics is no longer constrained by hardware, but by the intelligence to perform tasks autonomously and safely.

AI for Science (AI4S)

AI is positioned as a fundamental scientific research tool capable of accelerating breakthroughs in key areas. The dedicated “AI for Science” zone at WAIC 2025 signals its institutionalization.

Examples, such as AI platforms that assist in discovering novel drugs, illustrate AI’s potential to accelerate research and development cycles. The active participation of pharmaceutical and materials science companies shows that AI4S is being integrated into corporate R&D pipelines to solve complex problems and create intellectual property. Generative AI is poised to transform every aspect of science, from drug discovery, to new materials, and understanding of complex biological systems. There is a lot of excitement around this idea. It does, however, require high quality training datasets and infrastructure for training large models.

Navigating the Chip Landscape

Faced with geopolitical and supply chain pressures, China’s AI chip strategy exhibits a dual approach: domestic substitution and exploration of new computing paradigms.

Domestic chip design companies are releasing new GPU products and are prioritizing optimization of “inference” rather than “training” scenarios. This is a pragmatic market strategy that focuses on capturing the largest market segment. Inference chips are critical for running AI applications, and represent a large market need as AI becomes more ubiquitous.

The “AI Supply Chain Resilience” forum highlights industry concerns about supply chain security and addresses issues such as hardware design, supply chain diversification, and the development of chiplet technologies. The ability to manufacture high end chips is becoming critical to national security. China is making huge investments in domestic chip manufacturing capabilities.

Overall, WAIC 2025 portrays a spirit of “pragmatic innovation” shaped by commercial needs and geopolitical realities. These technologies are less about abstract pursuits and more about solving concrete problems. These improvements may have a deeper impact on economic competitiveness and national technological security.

These parallel advancements—domestic chips, specialized models, robots, and research platforms—are components of a full-stack AI ecosystem with end-to-end capabilities. This ecosystem aims at supply chain independence, while being open to external technologies when possible. The ability to control the whole stack is important.

The AI-Driven Commercial Revolution

The focus has shifted from technology capabilities to ROI, showcasing the commercial value and competitive advantages generated by AI.

Smart Manufacturing & Future Industries

AI is driving a profound transformation in manufacturing, evolving from isolated pilot programs to a comprehensive restructuring. AI is becoming the “industrial brain” that optimizes supply chain collaboration, production planning, and energy consumption.

The case of Baosteel achieving cost savings through “industrial brain” solutions is a testament to AI’s ability to generate significant economic benefits. This aligns with the “industrial metaverse” concept where AI-driven optimization strategies are implemented in digital twins of physical production lines. We will continue to see adoption of AI across different processes, from design to manufacturing to supply chain, resulting in significant productivity improvements.

Reimagining Finance & Commerce

In FinTech and e-commerce, AI applications are shifting from front-end customer service to core back-end operations such as financial modeling, credit approval, and fraud detection.

The Fintech AI Forum highlights these shifts, as AI is being used for mission-critical functions. This shift confirms that AI systems have met the standards for reliability, stability, and security. Trust in AI systems is key to adoption, and will be even more important with systems making complex decisions.

Transforming Healthcare

AI applications are reaching scale and are obtaining approval from federal regulatory agencies for AI-enabled medical devices.

The approval of an AI diagnostic tool for early cancer screening marks a milestone in moving beyond academic research and pilot programs. This signals the potential for widespread commercialization of AI medical products and is helping AI move from a health management role, to a role as a serious treatment. AI is helping doctors be more efficient in how they diagnose and treat their patients; however, there is still the need to have doctors involved in the process.

Across these sectors, quantifiable ROI is the key metric for success at WAIC 2025. Companies that highlight economic value will attract more capital and customers.

AI adoption is most mature in sectors with state influence: heavy industry, finance and healthcare. The government incentivizes state-owned enterprises to adopt domestic AI technologies, creating a market for AI enterprises. This greatly reduces these AI companies’ innovation and market risks. This is a smart strategy that builds momentum in key areas of their economy.

Ecosystem Dynamics and Key Collaborations

Understanding China’s AI requires analyzing the interplay between industry giants, innovators, academia, and capital flows. WAIC 2025 provides a roadmap for this.

Powerhouses Pivot to Vertical Solutions

Tech giants such as Baidu, Alibaba, Tencent, and Huawei are shifting from providing general-purpose platforms to offering end-to-end, vertical solutions for specific industries.

Their exhibits highlight this shift in strategy, transitioning from AI “tool” vendors to partners selling business benefits. This makes it easier for their customers to adopt and use AI.

Specialized Innovators Emerge

To succeed, companies are specializing to build competitive advantages.

Companies such as “RoboForge” are creating specialized applications, know-how, and data insights. These companies are the key to maintaining a competitive and innovative ecology. The most important thing these companies do is that they are closer to the customer, and can leverage new learnings to create new applications.

Seamless Academia Pipelines

The connection between academia and industry is becoming more formalized. Academic research is being integrated into a pipeline from discovery to commercialization.

University-corporate partnerships are creating joint collaborations on large language models for specific industries. The Shanghai AI Laboratory is connecting basic research with commercialization by directing its models toward specific industry applications.

This matrix highlights “The Great Specialization,” with companies penetrating vertical markets. This maturity marks the end of “land grabs” and emphasizes talent strategies that combine AI with industry experience. This shows both that AI is becoming more mature and specialized, and that talent is becoming more specialized.

The government creates markets in key industries while enabling universities and entrepreneurs to add new talent to AI. They realize it takes both entrepreneurs and mature companies to really have sustainable impact.

This approach is a highly complex industry policy strategy, with large enterprises adding scale and entrepreneurs creating agility. This builds an AI ecosystem for resilience and robustness. The goal is to continue to build their strength in AI.

Governing the Intelligent Era

AI governance—setting rules and safeguards —is a vital dimension of WAIC 2025. In China, AI governance shapes national order and also encourages international influence.

Proactive Regulations

China is proactively building safety guidelines to ensure innovation, and to manage ethical risk.

The draft Basic Security Requirement for Generative AI has been circulated to encourage compliance. This ensures that AI tech stays on track with national values, while providing companies the ability to comply with globalized standards. The Chinese government wants to be proactive in developing regulations, rather than reactive, as it enables innovation.

International Dialogues

WAIC is an important platform for China to build international partnerships around AI.

The forum included speakers from the EU and ASEAN. The lack of significant commitment from the US points to China’s efforts to form alliances with Europe. They are trying to build relationships with countries outside the US sphere of influence.

This approach reflects a geopolitical standard-setting strategy. These standards are being promoted as a means to govern global commerce. China is trying to establish norms that can run alongside Western ideals. By creating AI governance, China is offering the world a separate way of governing AI.