| Industry Track | Definition and Core Value | Current Status and Trends | Representative Companies/Products (for reference only) |
| 🏢 Enterprise Services | Focus oncost reduction and efficiency improvement and business process automation, such as intelligent marketing, customer service, human resource management, financial analysis, etc. | KA (key accounts) and SMB (small and medium-sized businesses) markets are growing simultaneously; transitioning from “tools” to “partners“, emphasizing deep integration with existing business systems | Meituan, Yonyou Network, Baidu Smart Cloud, Kingdee International, Salesforce, Beisen (acquired Cool Academy) |
| 🎮 Gaming | AI-native gameplay design, NPC intelligent interaction, game content generation, enhancing immersion and gameplay diversity | Currently focused onefficiency tools; in the future, AI-driven content generation and interaction will be the core direction | Tencent Games, NetEase, Giant Network, Unity, Roblox |
| 🏥 Healthcare | Assisted diagnosis, drug development, personalized treatment, hospital management process optimization, etc. | Policy-driven industry intelligent transformation; multimodal capabilities are deeply applied, but with high requirements for accuracy and compliance | Yilian, ThinkMedical, Shukun Technology, NVIDIA Clara, Insitro |
| 📊 Marketing | Content generation, personalized recommendations, user insights, advertising optimization, addressing marketing efficiency pain points | Content generation is the most prominent scenario currently; AI is reshaping marketing processes, pursuing brand effectiveness | BlueFocus, HuiLiang Technology, TradeDesk, Jasper.ai |
| 📚 Education | Personalized tutoring, intelligent grading, virtual teachers, course content generation, achieving tailored education | Policies are actively promoting the integration of generative AI with education; AI technology has a natural fit with subject content | New Oriental, Homework Help, iFLYTEK (Xingchen Agent platform), Duolingo Max, Khan Academy |
| 📈 Finance | Intelligent investment advisory, risk control, quantitative trading, compliance monitoring, customer service, etc. | Compliance and timeliness drive financial models to marginal deployment and lightweight iteration | iFinD, Hang Seng Electronics, China Merchants Bank, Ping An Group, BloombergGPT |
| 🤖 Office and Efficiency | Content creation (text/PPT), meeting minutes, schedule management, email processing, code generation, etc. | Text processing, presentations, data processing, and task management are four major components of AI office software | Kingsoft Office, Microsoft Office Copilot, Notion AI, Zoom IQ, GitHub Copilot |
| 🚀 Developer Tools and Platforms | Providing low-code/no-code AI Agent building, testing, and deployment platforms and frameworks, lowering development barriers | Major model vendors and cloud service providers are focusing on this area; promoting the standardization and scalable application of AI Agents | iFLYTEK (Xingchen Agent platform), Baidu Wenxin large model + Smart Cloud, AutoGLM2.0, LangChain, LlamaIndex |
🔍 Comprehensive Investment and Financing Observation
Investment and financing activities in the AI Agent field in China have shownexplosive growth in the past two years. As of July 2025, the financing amount for the year has exceeded8.2 billion yuan, and it is expected to exceed15 billion yuan for the entire year. The financing stages have gradually shifted from early exploration toconcentrated mid-to-late stage rounds, with a significant head effect.
Investment shows obvious regional and subject characteristics: forming three core gathering areas: Beijing (policy-driven), Shanghai (highly internationalized), Hangzhou (ecological collaborative). State-owned background funds (such as the Beijing Artificial Intelligence Industry Investment Fund) focus more on underlying technology “bottleneck” areas; industrial capital (such as Alibaba, Tencent) binds ecology through “technology + scenarios”; while VCs focus more on application layers and clearly commercialized tracks.
Mergers and acquisitions cases are also increasing, such as Beisen’s acquisition of Cool Academy, and it is expected that from 2025 to 2027, there will be a wave of mergers worth hundreds of billions.
📈 Market Prospects and Driving Factors
- • Strong policy drive: The State Council issued the “Opinions on Deepening the Implementation of the ‘Artificial Intelligence +’ Action”, aiming for widespread and deep integration of artificial intelligence with six key areas by 2027, with the penetration rate of new generation intelligent terminals and agents exceeding70%. This provides strong momentum for the implementation of AI Agents across various industries.
- • Market scale and growth: According to Fortune Business Insights, the market size of AI Agents in China is expected to soar from 57.4 billion yuan in 2023 to3.3009 trillion yuan by 2028, with a compound annual growth rate exceeding 100%. The global market is also considerable, expected to grow from 5.29 billion USD in 2024 to47.1 billion USD by 2030.
- • Technological evolution and commercialization breakthroughs: The enhancement of multimodal interaction capabilities, Agentic AI (intelligent agent AI), and multi-agent collaboration are becoming important trends. At the same time, more and more companies (such as Meituan, Kuaishou) have disclosed AI-related revenues in their financial reports, and commercialization is accelerating.
⚠️ Investment Risks and Challenges
Investing in the AI Agent track also requires attention to the following risks:
- 1. Technological progress not meeting expectations: The iteration of large model technology slows down or encounters bottlenecks.
- 2. Commercialization not meeting expectations: Slow market acceptance, unclear profit models.
- 3. High computing costs: Significant cost pressure for model training and inference.
- 4. Data security and compliance risks: Increasingly strict requirements for data privacy protection and algorithm compliance.
- 5. Intensifying industry competition: Tech giants and startups competing in the same arena, some tracks may quickly become red oceans.
Scan for more exciting content