At the 2026 World Artificial Intelligence Conference (WAIC) and High-Level Meeting on Global AI Governance, held in Shanghai from July 17 to 20, Ant Group Co., Ltd. CEO Han Xinyi shared his insights on how corporate AI can progress from initial deployment to delivering deep, tangible value.
Core Principles for Enterprise AI
Han Xinyi outlined three key considerations. He emphasized that the intelligence of enterprise AI should not be limited to knowledge accumulation but must also encompass the sedimentation of practical experience. He pointed out that general-purpose large language models trained on public data may not truly grasp a specific company's operational logic. Genuine enterprise-grade AI intelligence, he argued, needs to be rooted in real business processes, continuously accumulating proprietary data and experience, autonomously learning and identifying inefficiencies through actual operations to develop judgment tailored to the company's unique business logic. He termed this direction "Experience AI."
Moving Beyond Tool-Level Deployment
Han noted that many current enterprise AI agent products remain at the tool deployment stage, assisting employees with tasks like searching information, writing emails, summarizing meetings, or reviewing contracts. However, merely focusing on superficial individual productivity gains may fail to address organizational issues and could even become constrained by existing procedural inertia. He explained that many current corporate workflows are designed for "people," not for "Agents." Traditional processes involving step-by-step information relay and confirmation often lead to a scenario where "each link performs its best, but when all links are connected, no one takes responsibility for the final outcome."
The Path to Deep Integration and Organizational Intelligence
According to Han, the true value of AI agents lies not just in helping employees complete tasks faster, but in serving as intelligent advisors to restructure processes. They should be capable of discovering and diagnosing problems, assisting employees in making trade-off judgments, and finding solutions that balance business advancement with risk control. He believes the path for enterprise AI to achieve deep integration is to build "organizational intelligence."
Han envisions a future where every enterprise will possess a vast number of digital employees, each human employee will have their own AI agent partners, and these agents will form new collaborative networks, constituting an AI organizational structure. He suggested that companies should first deploy AI agents into every key process to accumulate proprietary data and experience, as without this foundation, there can be no enterprise-specific intelligence. Subsequently, a trustworthy data foundation should be established, utilizing technologies like blockchain and privacy computing to ensure data authenticity, traceability, and secure flow throughout the entire process, ultimately forming a continuously iterative closed loop of experience.
"The era of the AI economy has arrived," Han stated. He concluded that AI will only replace standardized and repetitive work, not human judgment and creativity. AI will amplify and reinforce the corporate experience accumulated by people, making organizations more intelligent and efficient, and achieving unity between business growth and employee development. "This is the true meaning of technology serving humanity."
During the forum, Ant Digital Technologies officially launched its "Intelligent Agent Super Factory" solution, aimed at helping enterprises build new production systems for the AI era.
Comments