At the 2026 Inclusion· Bund Summit forum on "New Paradigms of AI Economic Growth and Reconstruction of Social Value" held on the afternoon of September 9, Ant Group CEO Han Xinyi stated that the AI economy cannot rely solely on capital investment, and AI applications must not stop at improving existing operational efficiency. He emphasized that enterprises need to transition from cost reduction and efficiency enhancement to value creation, using AI to generate new supply, stimulate new demand, and empower individual workers and small and medium-sized enterprises with the capability and opportunity to participate in value creation and share in the dividends of AI-driven growth.
"There is no doubt that AI is one of the most significant variables in today's economic and social landscape," said Han. Investment in foundational infrastructure, from chips and computing power to data centers, continues to rise, and AI is increasingly entering production and daily life scenarios through large models and intelligent agents. However, from an economic and social development perspective, a crucial question still needs to be addressed: How can AI lead to more sustainable economic growth and broader social value?
Currently, the world is witnessing an unprecedented wave of AI infrastructure investment. Capital expenditure planned by just four companies—Amazon, Microsoft, Google, and Meta—is projected to exceed a combined $700 billion in 2026. Han pointed out that the stimulating effect of such capital expenditure on the economy is temporary; only when these investments are converted into sustained future output and value can they potentially drive genuine long-term economic growth.
Han warned against an AI economy built purely on capital, stressing that applications must go beyond enhancing existing tasks. If AI remains confined to simple efficiency gains and substitution without creating incremental demand, it could lead to new problems such as supply-demand imbalance and mismatched returns on investment. He argued that "businesses must move from cost reduction to creating new value, generating fresh supply and activating new demand to foster a richer intelligent economic ecosystem."
He noted that promising pathways and trends are emerging around these new demands and business models. On one hand, when AI can organize models, data, tools, corporate knowledge, and human expertise, services that were previously too costly, reliant on scarce experts, or hard to scale are entering the market to meet previously unmet needs. Take healthcare as an example: Ant Group's "A Fu" application helps users understand medical reports, compile long-term health records, offer health advice, and connect them with suitable hospitals and doctors. Currently, 55% of A Fu users come from third-tier cities and below, nearly 30% are elderly, and about one-sixth of consultations occur between midnight and 6 a.m. Han believes AI has the potential to significantly improve access to high-quality, specialized medical services.
On the other hand, a new intention-centric intelligent agent economy is beginning to unlock fresh demand. Han explained that traditional commerce is often "traffic-centric," with merchants relying on large traffic platforms, but as traffic acquisition costs rise, marginal returns diminish. In the agent era, the core of business shifts towards understanding user "intentions"—accurately interpreting needs and ensuring suitable products and services appear when a user demands them. If every user has their own personal life assistant and every merchant has a dedicated business agent, these agents can connect based on actual needs and complete transactions through agentic payment systems. Han suggested this new commercial loop could generate incremental demand vastly different from what the mobile internet era produced.
Once new demand is created, Han posed another essential question: How can the growth generated by AI translate into employment and income opportunities for individual workers and SMEs, creating a new cycle of "investment-income-consumption"? Realizing this cycle requires a more balanced distribution between capital and labor, supported by robust distribution and social security mechanisms. He addressed the employment issue, noting some propose "human-reserved positions" in response to potential AI disruption. While acknowledging the rationale for roles involving ethical responsibilities, like judicial rulings, Han cautioned that broadly institutionalizing existing jobs through administrative means could sacrifice productivity. The more sustainable path lies in enhancing the ability of workers and SMEs to engage in new value creation and share AI's benefits. This involves companies distilling complex model capabilities into low-cost, user-friendly products that solve practical problems, lowering the entry barrier for AI usage. Simultaneously, it requires enhancing human skills through education and vocational training, leveraging judgment and creativity in human-machine collaboration. Han also stressed the need to refine secondary distribution mechanisms and establish fitting social safeguards to support workers through the transition, enabling them to re-enter the new economic cycle.
"AI's value will ultimately be judged by real growth, and measured by the sense of gain among a broader population," Han concluded.