Technology - The competition in AI large models will continue into 2026, with companies like OpenAI, Google, and DeepMind releasing larger or more efficient models [2] - AI models are making progress in spatial understanding, aiming to develop capabilities in semantics, physics, geometry, and dynamic interactions [2] - By 2026, 40% of enterprise applications are expected to embed task-oriented AI agents, a significant increase from less than 5% in 2025 [2] Industry - The combination of digital twins and AI agents is reshaping product design processes, marking a strategic opportunity for "smart manufacturing" [4] - By 2026, 40% of manufacturers with production scheduling systems are predicted to upgrade to AI-driven scheduling for autonomous resource management [4] - The shift from traditional manufacturing to "smart manufacturing" is expected to enhance market perception, product innovation, and international competitiveness for Chinese enterprises [4] Energy - The energy demand from global data centers is projected to double by 2030, reaching approximately 945 terawatt-hours, with AI being a major driver of this increase [5] - The global green AI data center market is expected to reach $67.6 billion by 2026 and grow to about $123 billion by 2035 [5] - China is focusing on optimizing supply capabilities and green low-carbon initiatives to support sustainable AI development [5] Governance - 2026 is anticipated to be a pivotal year for the acceleration of global AI governance measures, shifting focus from ideological debates to compliance and cross-border collaboration [7] - The EU's AI Act, set to take effect in August 2026, will be the first comprehensive regulation of AI globally [7] - China's AI governance path is becoming clearer, with a focus on legal frameworks and ethical guidelines to promote healthy AI development [7] Overall Outlook - The future of global AI development will hinge not only on the strength of models but also on the integration of safety, compliance, energy consumption, and industry implementation into a cohesive system [8]

Technology - The competition in AI large models will continue into 2026, with companies like OpenAI, Google, and DeepMind releasing larger or more efficient models [2] - AI models are making progress in spatial understanding, aiming to develop capabilities in semantics, physics, geometry, and dynamic interactions [2] - By 2026, 40% of enterprise applications are expected to embed task-oriented AI agents, a significant increase from less than 5% in 2025 [2] Industry - The combination of digital twins and AI agents is reshaping product design processes, marking a strategic opportunity for "smart manufacturing" [4] - By 2026, 40% of manufacturers with production scheduling systems are predicted to upgrade to AI-driven scheduling for autonomous resource management [4] - The shift from traditional manufacturing to "smart manufacturing" is expected to enhance market perception, product innovation, and international competitiveness for Chinese enterprises [4] Energy - The energy pressure from AI applications will remain high, with global data center electricity demand expected to double by 2030, reaching approximately 945 terawatt-hours [5] - The global green AI data center market is projected to reach $67.6 billion by 2026 and could grow to about $123 billion by 2035 [5] - China is focusing on sustainable resource and engineering systems to support AI development through supply capacity and green low-carbon initiatives [5] Governance - 2026 is anticipated to be a critical year for the acceleration of global AI governance measures, shifting focus from conceptual debates to compliance and cross-border collaboration [7] - The EU's AI Act, set to be implemented in phases starting August 2026, will be the first comprehensive regulation of AI globally [7] - China's AI governance path is becoming clearer, with a focus on improving legal frameworks and ethical guidelines to promote healthy AI development [7]

新华社北京1月20日电 特稿｜展望全球人工智能2026年演进新局 新华社记者孙晶 站在2026年的起点，展望全球人工智能（AI）发展，技术、产业、能源、治理多重变量交织，将共同 塑造这一关键年份。 相关机构预测，越来越多的顶尖AI企业将聚焦提升大模型推理能力与智能体执行任务能力，推动AI 从"会生成"向"会规划、会行动"进化。大量企业应用将嵌入任务型AI智能体。 与技术突破相伴的则是能源压力，全球数据中心耗电量将持续高企。治理层面，预计各国治理措施将加 速落地。 产业："智能制造"迎来机遇期 在产业界，数字孪生与AI智能体结合正在重塑产品设计流程，"智能制造"迎来战略机遇期。 美国国际数据公司预测，2026年，40%配备生产调度系统的制造商将升级采用AI驱动的生产排程，实现 生产资源管理的自主化运行；到2028年，全球头部1000家制造企业中将有65%把智能体与设计、仿真工 具结合，用于持续验证设计变更与配置方案。 美国液态人工智能公司联合创始人兼首席执行官拉明·哈萨尼认为，今年将是"主动智能体"之年。他 说，目前大多数AI助手等都是"反应式智能体"，但当AI在设备上快速运行且始终在线时，它可以主动为 人类工 ...