Visitors examine an exploded-view display of a humanoid robot at the 2026 World Robot Conference. Photo: Lyu Jiazuo/CSST

The rise of the platform economy has made information, traffic, and control over user access increasingly central to market competition. Over the past two decades, search engines, e-commerce platforms, social media, and content recommendation systems have brought vast supplies of goods and services within reach of finite consumer demand, markedly lowering search costs and placing attention at the heart of competition. Businesses that control these gateways, secure favorable recommendation rankings, and keep users engaged longer are better placed to attract traffic, earn advertising revenue, and turn visits into purchases. Attention has thus become a scarce resource in digital markets, with platforms serving as key intermediaries connecting supply and demand, directing traffic, and shaping consumer choices.

China’s AI development has entered a new stage, moving from technological breakthroughs to large-scale application—a shift reflected in a series of policy initiatives. In 2025, the State Council issued the Opinions on Deeply Implementing the “AI Plus” Initiative, calling for faster development of a smart economy and intelligent society built on human–machine collaboration, cross-sector integration, and joint creation and sharing. The document set a target of over 70% adoption for applications such as new-generation smart terminals and AI agents by 2027. The 2026 Government Work Report further called for “creating new forms of smart economy,” accelerating the adoption of new-generation smart terminals and AI agents and expanding commercial AI applications across key industries. In May, the Implementation Opinions on the Standardized Application and Innovative Development of AI Agents set out measures to guide their use, support innovation, and govern security risks. Together, these policies point to a growing economic role for AI agents, which are becoming important gateways to—and integral components of—this emerging smart economy.

From attracting traffic to understanding demand

The attention economy emerged from a mismatch between abundant information and the limited human capacity to process it. Where consumers in traditional markets struggled to find enough information, those using digital platforms face information overload, excessive choice, and the considerable effort of evaluating their options. Platforms reorganize information through recommendations, rankings, and advertising to match supply with demand more efficiently. Yet the benefits have limits: a wider selection also leaves consumers with more options to sift through. Businesses may become preoccupied with impressions and clicks, channeling resources into promotional packaging, emotional appeals, and efforts to generate traffic at the expense of sustained improvements in product quality and service delivery. Rising costs of attracting traffic can also push small and medium-sized enterprises (SMEs) to the market’s margins.

With the emergence of AI agents, visibility is giving way to a further competitive consideration: how accurately users’ needs can be understood. When consumers delegate travel planning, purchasing, financial planning, or health management to AI agents, they convey more than the fragmentary signals captured by clicks and brief visits. Their requests specify goals, budgets, preferences, deadlines, and acceptable levels of risk, giving agents a clearer account of what they want to achieve. Competition increasingly turns on the ability to capture these requirements and interpret the intentions behind them, shifting the emphasis from attracting traffic to understanding demand.

Businesses will therefore need to consider how AI agents evaluate their offerings alongside how those offerings appeal to consumers. Page design, promotional language, and advertising placement have long exerted considerable influence on purchasing decisions. As agents take a greater role, their selections will depend directly on standardized product specifications, clear service terms, credible quality certification, and a verifiable record of who bears responsibility for after-sales service. These qualities will carry increasing competitive weight as businesses seek to make their products and services easy to understand, straightforward to compare, and worthy of trust.

From platform matching to reliable fulfillment

The platform economy primarily brings buyers and sellers together, helping them find each other more quickly. E-commerce recommendations, ride-hailing dispatch, content distribution, and advertising placement all aim to connect supply with demand more efficiently. The intention economy extends this role to fulfillment—ensuring that transactions are completed and commitments honored. AI agents can identify suitable options, compare alternatives, negotiate prices, make payments, track delivery, and provide feedback, turning a user’s stated needs into a coordinated sequence of actions from initial search to completed transaction.

This change will reshape the competitive advantages of enterprises. Success in the attention economy has depended chiefly on attracting traffic, building brand visibility, and managing a presence on digital platforms. As the intention economy develops, reliable fulfillment will carry greater weight. AI agents will keep ongoing records of how well businesses deliver, how stable their prices remain, how they handle after-sales service, and whether they honor their commitments. These records will inform subsequent purchasing decisions, giving each transaction a bearing on future business. Beyond winning individual clicks or purchases, firms will compete to remain among the suppliers that AI agents trust and select over time.

SMEs may also find new opportunities as the basis of competition shifts. Many struggle to afford the advertising and platform operating costs required to attract traffic. In the intention economy, clear product standards, a reliable fulfillment record, and distinctive services could help AI agents identify these businesses and recommend them to users whose needs they meet. Provided that market interfaces are sufficiently open and data rules transparent, agents could lower barriers sustained by monopolistic control over traffic, allowing quality, efficiency, and trustworthy service to play a greater role in determining which businesses succeed.

From platform rankings to AI gateways

The transition to the intention economy will transform the power of intermediaries without necessarily diminishing it. Platforms have shaped user choices through their homepages, search rankings, recommendation algorithms, and advertising systems. As AI agents gain ground, the gateways through which users access them, model interfaces, default agents, payment protocols, identity authentication, and data authorization mechanisms may form a new layer of market infrastructure. Control over the underlying channels connecting agents to the market could confer considerable influence over how users express their intentions, how agents interpret them, and what actions follow—making these channels a potential source of market power.

The intention economy could therefore empower consumers while creating new forms of dependence on agent providers. AI agents can help users search and compare offers across platforms, reduce information asymmetry, and strengthen their bargaining power. Yet if a few providers control high-quality agents, key data interfaces, and default service gateways, model preferences, commercial partnerships, and interface restrictions could narrow users’ choices. What appears to be a “personal agent” acting on the user’s behalf could itself become another platform gateway.

This requires market regulation to extend from traditional platform governance to the governance of AI agents. Alongside established concerns about monopolistic control over traffic, algorithmic recommendations, and data abuse, regulators must address control over agent gateways, closed interfaces, manipulation of default options, and barriers to cross-platform interoperability. The stakes are particularly high in consumer finance, online healthcare, AI-assisted procurement, and automated price negotiations. Once agents are authorized to act on users’ behalf, their decisions and actions directly affect users’ rights and the orderly functioning of markets.

Building institutions to support the intention economy

The healthy development of the intention economy requires the support of appropriate market institutions. First, transaction information should be standardized and made machine-readable, enabling AI agents to accurately interpret and compare information on product quality, pricing rules, service terms, after-sales responsibility, and credit records. Second, a system for identifying agents and managing their permissions should define the limits of their authority, require records of their actions, and establish clear lines of responsibility. This would help prevent unauthorized actions and ensure that liability can be assigned when problems arise. Third, rules on interface openness and interoperability should be strengthened to prevent a handful of platforms from erecting new market barriers through closed protocols and default agents.

Value alignment also deserves close attention. AI agents pursue objectives expressed in machine-readable terms, while people’s actual intentions often involve complex requirements concerning quality, safety, fairness, privacy, and long-term interests. An agent asked to find the “lowest price” must still account for quality risks. Similarly, an instruction to “maximize conversion” cannot justify violating consumer rights or undermining fair competition. Developing the intention economy therefore calls for stronger algorithmic capabilities supported by clearer rules on responsibility, more transparent auditing of how agents operate, and more robust risk assessment.

The transition from the attention economy to the intention economy marks a significant shift in digital markets, from making information visible to making users’ intentions actionable. The attention economy changed how people discover goods and information. The intention economy will reshape how their needs are understood, transactions are carried out, and value is created. As this new smart economy takes shape, AI agents should be developed and deployed with openness, interoperability, reliable fulfillment, and value alignment as guiding principles. These foundations can help redirect market competition toward value creation, giving businesses stronger incentives to meet users’ needs and supporting high-quality development and the construction of a modern industrial system.

Chai Hongfeng is a professor from the Institute of Financial Technology at Fudan University. Chen Ruoxin is from the College of Management and Economics at Tianjin University.