The AI agent era is speeding up, and listed companies are actively planning and positioning themselves.

“Future AI agents could outnumber humans, and humans will live in a world with hundreds of millions, or even tens of billions of AI agents.” Meta CEO Mark Zuckerberg predicted this in an interview program in July this year. A report from Bank of America also pointed out that AI evolution is on the eve of a third wave—agentic AI, i.e., intelligent agent AI, is about to arrive.

In the second half of this year, AI Agents have been all the rage, becoming a hot area that tech giants such as Microsoft, Apple, Google, and OpenAI are throwing their efforts behind. In China’s domestic market, large-model vendors, internet companies, and listed companies are all actively laying out their plans for AI Agents.

As more and more companies roll out AI Agent products and ecosystems, industry insiders expect 2025 to be the breakout year for AI Agents. In its research report, Galaxy Securities projected that the scale of China’s AI Agent market will surge to 852 billion yuan by 2028, with a compound annual growth rate of 72.7%.

Vendors kick off the AI agent battle

AI Agent, which stands for an artificial intelligence agent and is also called an “AI intelligent agent,” is a system driven by large language models as its brain. It has capabilities for autonomous understanding, perception, planning, memory, and using tools, enabling it to automatically execute and complete complex tasks. Unlike traditional AI, AI Agents can achieve the ability to gradually complete a given goal through independent thinking and by calling tools.

For a simple example: if a user wants to go out for a meal, a large language model can provide dining location suggestions and prompts about relevant store information. But an AI Agent can do more than recommend where to eat—it can also search for cuisines and restaurants based on the user’s budget, carry out the booking, add the itinerary to the calendar, and send itinerary reminders.

Since the second half of this year, AI Agents have been highly sought after, and major tech players such as Microsoft, Apple, Google, OpenAI, and Anthropic have all announced relevant progress.

In June, Apple demonstrated its latest AI results, Apple Intelligence, at its developer conference. In November, Microsoft released more than 10 commercial AI Agents at the “Microsoft ignite 2024” conference. Google followed closely, also announcing a push to promote commercial AI Agents, releasing a series of incentive activities and products—and additionally, specifically rolling out one of the few commercial AI Agent markets worldwide. OpenAI, meanwhile, plans to release a new AI Agent product code-named “Operator” in January 2025.

In the domestic market, vendors represented by Zhipu (智谱) are also focusing their efforts on AI Agents, and a quiet battle over “AI Agents” is already underway.

On October 25, Zhipu released AutoGLM, an intelligent agent that can read voice commands to understand user intent, simulate human behavior, and automatically complete actions such as ordering takeout, booking flights, and booking hotels. On November 29, Zhipu rolled out an upgraded AutoGLM version that supports autonomous execution of long-step tasks of more than 50 steps, and can flexibly switch between different apps during task execution.

Zhipu believes that AI large models are moving from Chat to Act. In the future, a unified set of AI intelligent agents will operate various hardware devices to improve human-computer interaction efficiency. With computing power continuing to rise, models and terminal-cloud collaborative architectures that are tailored for AI-native devices will gradually emerge. Various smart devices such as phones, PCs, cars, glasses, and home appliances are rapidly coming to market.‌

Besides Zhipu, companies including ByteDance, Baidu, Tencent, and Alibaba are also deploying AIAgent, including ByteDance’s Coze Agent platform, Baidu’s Qianfan AgentBuilder agent development tool, and TianGong skyAgents from Kunlun Wanwei.

Multiple listed companies roll out AI Agents

While domestic vendors initiate the AI Agent battle, multiple listed companies in China’s A-share market are also actively laying out AI Agents.

For example, iFlytek (科大讯飞) has launched AI assistants for fields such as education, healthcare, justice, and government services. Jincai Interconnection’s “Xin Zhiyue Taxation & Finance Large Model” adopts an open architecture design. By combining the company’s taxation and finance products and an open platform, it builds a taxation and finance AIAgent intelligent agent.

The AI Agent development and operating platform from Cxjin Cai (中科金财) provides functions such as agent creation, multi-basis model invocation, and workflow definition. It can automatically route and dispatch the most suitable large model based on industry scenario needs and complete agent creation. The company states that its Cxjin Cai AI Agent achieves multi-intent understanding: it can automatically execute multiple tasks based on user instructions or contextual information, and after learning user preferences through understanding, it can provide personalized services, enabling highly human-like end-to-end natural language interaction.

Xingguodou Holdings’ subsidiary Shanghai Shiershi is dedicated to research and development of General Purpose AI Agent (general-purpose intelligent agent) technology, carrying out research on multimodal AI technologies and products.

XinkaiPu launched a campus life AI product named “Xiaomei (小美) Classmate” in October this year, built on the PanGu large model and Tongyi Qianwen large model.

On December 9, when CSGO (中科创达) was asked by investors whether it has already deployed AI Agents, it said the company has been pushing product and technical development in the area of AI Agents. For example, its Rubik Avatar is an integrated and innovative product that combines multiple technologies, including AI Agents. It uses AI technologies including machine learning, natural language processing, and computer vision. Based on on-device intelligence and naked-eye 3D technology, and on Qualcomm’s high-computing-power chip platform, it uses the Kanzi 3D rendering engine to render 3D graphics and images in real time.

Nangxing Shares stated in its recent reply to investor questions that its subsidiary Only Network is based on cutting-edge AI large model technology. It has carried out deep cooperation with industry-leading AI companies such as Microsoft to provide efficient and intelligent AIGC application products and solutions for industry customers. XiaoLu is an AIGC-based AI assistant that supports intelligent chat and knowledge-base Q&A.

In a reply to investors on December 17, Qute Intelligence said that the goal of its Qute AI Agents 2.0 version is to become a universal digital-intelligence tool and helper, helping traditional enterprises with digital-intelligence transformation and upgrading. At present, Qute AI Agents 2.0 is still under development.

2025 may bring a commercial breakthrough for AI Agents

For AI Agents, industry generally believes that 2025 will be the year when AI Agents explode.

A research report from China Merchants Securities (山西证券) points out that Agents are expected to be deployed in volume in 2025, driving up demand for compute and further expanding demand for investment in AI infrastructure.

Debon Securities also predicted that by 2025, AI Agents may enter a breakout period, and the terminal market will make a qualitative leap. As the Agent network gradually takes shape, higher market penetration will help construct a good closed-loop business model for the entire AI ecosystem.

“Right now, tech giants such as Apple, Google, and OpenAI have already viewed Agents as one of the key focuses for 2025. 2025 may become the first year of Agent breakout, thereby driving the rollout of AI applications.” said Zong Jianshu (宗建树), an analyst at Changjiang Securities.

In the view of Tang Fangxin, head of government and enterprise business at Wanxing Technology, AI Agent development is currently in the transition stage from “experiment” to “application,” and has not yet reached maturity. However, with continued improvement in computing power and progress in machine learning and natural language processing technologies, AI Agents’ capabilities in understanding user intent, providing personalized services, and executing complex tasks are continuously improving. At present, AI Agents have been applied in multiple fields such as customer service, smart home, and personal assistants, and are expanding into more industries and scenarios.

Tang Fangxin believes that AI Agents will continue to evolve and gradually move from being “usable” to being “really convenient.” They have already landed in some commercial scenarios—for example, the interactive digital human business launched by Wanxing Technology can largely solve welcome and presentation reception scenarios in exhibition halls, smart demonstrations, speech and reporting, and corporate guest welcoming. It introduces service content for specific scenarios to audiences and consumers through AI digital humans,革新 (innovates) traditional interaction methods with AI, and empowers improvements to audience experience.

However, Tang Fangxin pointed out that the introduction of AI Agents involves the collection and processing of large amounts of user data. Therefore, safety and privacy are crucial and may trigger potential privacy concerns such as data misuse and privacy leakage. Moreover, AI Agents must rely on local compute and large models, which imposes higher requirements on technical maturity, vendor integration and implementation capabilities, and operational and maintenance service capabilities. Currently, efforts should shift from penetrating a single scenario to related scenarios, continuously accumulate mature paths and solution experience, and at the same time strengthen safety and privacy regulation while exploring and embracing it with an open mindset.

“From ‘training’ to ‘reasoning,’ AI intelligent agents will become the inevitable path to the era of general artificial intelligence.” said a Galaxy Securities research report. With large models rapidly iterating and upgrading, AI intelligent agents based on large-model technology have entered a phase of rapid development. The rise of AI intelligent agents is reshaping the AI industry chain and bringing new investment opportunities. It projects that by 2028, the market size of China’s AI agent market will surge to 852 billion yuan, with a compound annual growth rate of 72.7%. The AI intelligent agent industry chain is a diversified and highly coordinated ecosystem, with broad future market space.