Workflow
无人/智能驾驶
icon
Search documents
加速与应用——2025年全球人工智能技术、政策、产业与投融资趋势全景洞察报告-国泰海通
Sou Hu Cai Jing· 2026-02-01 11:00
Group 1: Global AI Industry Overview - The global AI industry in 2025 shows a differentiated pattern of "China's recovery and overseas cautious restructuring," with multi-dimensional collaboration in capital, policy, technology, and industry driving AI from "technological illusion" to a new stage of "commercial evidence" [1] - In China, AI financing reached 3,180 events in 2025, with a significant increase in H2 compared to H1, focusing on AI software applications and embodied intelligence with clear ROI [1][14] - Overseas financing declined quarterly, with a shift back to foundational large models and computing infrastructure, and a slowdown in IPO activity [1] Group 2: Capital Market Dynamics - In China, the application service layer accounted for 69% of AI financing events, with hardware and software sectors showing significant growth in H2 [16][22] - Early-stage financing represented over 56% of total financing events in China, indicating a trend towards lighter, vertical-focused startups [25][29] - In contrast, overseas AI financing events totaled 1,637 in 2025, reflecting a shift from "frenzied pursuit" to a more selective and cautious investment approach [37][38] Group 3: Policy Evolution - Countries are moving towards a dual-track approach of "sovereign construction + scenario-driven" policies, with China promoting full-scene value realization through "AI +" initiatives [1] - The U.S. is deepening its strategy of "internal innovation and external binding," focusing on exporting comprehensive technology standards [1] - European countries, the UK, Japan, and South Korea are emphasizing computing sovereignty, while Southeast Asia and the Middle East are accelerating the establishment of autonomous ecosystems and data hubs [1] Group 4: Industry Chain Innovation - The AI industry chain is characterized by "system-level energy efficiency competition," with large models focusing on deep reasoning and native multi-modal capabilities [2] - The AI chip competition has shifted to a three-dimensional game of "efficiency - ecology - customization," with edge computing enhancing "end-edge-cloud" collaboration [2] - Significant breakthroughs in engineering delivery have been achieved in robotics, autonomous driving, and healthcare applications, marking a transition to industrial-level evolution for humanoid robots [2] Group 5: Strategic Differentiation Among Major Players - Overseas giants like NVIDIA are focusing on reconstructing intelligent agent paradigms and computing infrastructure, while Google and Microsoft are enhancing their differentiated advantages [2] - Chinese companies such as Huawei, Baidu, and ByteDance are deepening their focus on vertical scenarios and open-source ecosystems, building core competitiveness through breakthroughs in computing power and ecosystem empowerment [2][10]