The 2026 World Artificial Intelligence Conference was not only a stage for emerging AI companies to unveil new technologies but also a crucial window for major internet companies to demonstrate their AI transformation achievements.

At the conference, internet giants such as Baidu, Alibaba, and Tencent showcased Agent tools, AI hardware, and AI cloud service platforms tailored to the needs of C-end users. However, notably, JD, another To C internet company, focused on demonstrating To B technologies such as head-mounted data collection devices, home showrooms, and inventory robots.

In response, Duan Nan, Vice President of JD Group and Deputy Dean of JD Explore Academy, stated, 'JD plans to establish the world's largest physical world operations center. We will focus on capabilities such as embodied intelligence and the multimodal understanding, reasoning, generation, and real-time interaction required for embodied intelligence to deploy the matrix of JD's JoyAI foundational models.'

As the wave of AI technology approaches, JD's alternative focus on To B technologies indicates its desire to align with its asset-heavy supply chain advantages for AI-driven transformation. This also suggests that JD may embark on an AI development path distinct from those of Baidu, Alibaba, and Tencent.

JD: Protected and Constrained by Its Supply Chain

For a long time, '3C e-commerce' has been JD's most distinctive label. However, when looking through the essence of its business, e-commerce is merely JD's gateway to consumers, with the supply chain being the core foundation supporting its business model.

On the eve of the 2025 618 shopping festival, Liu Qiangdong, founder and chairman of JD Group, stated, 'We are not a so-called diversified company. As you can see, all of JD Group's businesses revolve around the supply chain. I never engage in anything unrelated to the supply chain. Currently, 100% of JD's businesses are centered around the supply chain.'

In the e-commerce sector, JD has built an asset-heavy infrastructure integrating warehousing and distribution. Official data shows that as of now, JD Logistics operates over 3,600 logistics warehouses with a total management area exceeding 34 million square meters. Based on this, 95% of JD's self-operated orders can be fulfilled within 24 hours, and 'minute-level' delivery is even possible in nearly a hundred cities.

It can be said that JD's core competitiveness in e-commerce primarily lies in constructing a closed loop of genuine product quality, fulfillment certainty, and high-value users, with the 'self-operated + self-built logistics' super supply chain as its foundation.

Although in recent years, Pinduoduo has risen rapidly with its low-priced white-label products, it has not shaken JD's position in the high-quality e-commerce market due to significant shortcomings in product quality control, logistics, and fulfillment.

In recent years, JD's foray into the food delivery business has also revolved around the supply chain. Liu Qiangdong said, 'Everyone sees our competition with Meituan in food delivery, but the logic behind it is the fresh produce supply chain—that's what I truly want. Selling meals on the front end will never be profitable; I can make money through the supply chain.'

It is reported that in July 2025, JD launched a joint venture quality catering production platform called Qixian Kitchen, opening up its supply chain and operational capabilities to the catering industry to empower high-quality catering merchants and provide consumers with more quality-assured food delivery services.

Facts have proven that in-depth layout (layout) in the upstream industrial chain has indeed endowed JD's food delivery business with significant market competitiveness. Official data shows that each Qixian Kitchen outlet achieves an average daily order volume of over 500 orders three months after opening. On its first anniversary, JD's food delivery service had accumulated 240 million ordering users, with a market share exceeding 15%.

However, the drawbacks of a deep focus on the supply chain are also evident. Due to its heavy business model, JD's profit margins are significantly lower than those of asset-light internet companies. Financial reports show that in 2025, JD's net profit margin was only 1.5%. During the same period, Pinduoduo's net profit margin reached as high as 23.01%.

As Duan Yongping, founder of BBK and investor, once said, 'A business that doesn't make money is useless, no matter how much revenue it generates.' While betting on asset-heavy supply chains has provided the company with a relatively solid commercial moat, JD has not become a focal point in the capital markets due to its weak profitability. Since 2021, JD's U.S.-listed stock price has been on a downward trajectory, with cumulative declines exceeding 70%.

Now, as AI technology gradually matures, the internet industry has reached a critical juncture for a new round of transformation, with many major companies striving to seize C-end entry points.

In contrast, JD finds itself trapped in the supply chain 'cage' it has built.

JD Avoids Competing for Entry Points and Seeks to Revalue Its Supply Chain Through AI

Since 2026, as Agent technology has gradually matured, internet companies have been integrating their ecosystems to create AI assistants capable of executing tasks with a single sentence, hoping to secure their place in the AI era.

For example, in January 2026, Alibaba officially announced that its Qianwen App would be fully integrated with Alibaba ecosystem businesses such as Taobao, Alipay, and Fliggy, enabling AI-powered shopping functions like ordering food delivery, shopping, and booking flights.

Similarly, in June 2026, WeChat also began internal testing of WeChat AI. Developers can integrate mini-programs into the WeChat AI ecosystem. Based on WeChat AI, users can control a vast array of WeChat mini-programs with a single sentence using natural language.

In contrast, unlike internet giants such as Alibaba and Tencent, which possess rich ecological resources, JD has not actively pursued the deployment of general-purpose Agent entry points. Instead, it has chosen to open up its ecosystem to external platforms.

After WeChat AI began internal testing, JD announced that it would be among the first to integrate, implementing WeChat AI Agents across its e-commerce, food delivery, and logistics businesses. In mid-July 2026, JD's AI Agent also completed ecosystem integration with Yuanbao. Users can directly view information on JD's full range of products and place orders within Yuanbao.

While integrating with external AI assistants helps JD quickly seize the AI opportunity, its reliance on the 3C e-commerce entry point, which is crucial for its survival, may be weakened.

As AI assistants gradually become the new entry point in the mobile internet industry, users will primarily perform tasks through Agents, potentially reducing the frequency of opening third-party apps. At that point, JD will only serve a fulfillment role, losing its traffic entry point and user reach capabilities.

In light of this, in addition to actively integrating with external Agent platforms, JD has also increased its focus on To B-oriented AI technologies, leveraging its strategic advantages.

It is reported that JD's AI technology roadmap consists of three steps: 'digital intelligence, attached intelligence, and embodied intelligence.' First, transform JD's existing businesses through digital intelligence, then compete for household and smart hardware entry points through attached intelligence, and finally achieve embodied intelligence, enabling robots to perceive, understand, decide, and execute tasks in the physical space.

Overall, JD has embarked on an AI path that emphasizes 'heavy scenarios, heavy supply chains, and heavy deployment.' Rather than enthusiastically competing with large model companies on parameters and entry points, JD is committed to creating tangible value for AI technologies in operational, warehousing, service, and other scenarios.

As Duan Nan said, 'JD has rich business scenarios in retail, logistics, health, and industry, providing unique conditions for model evaluation.' Unlike traditional internet companies that operate with light assets, JD, with its vast asset-heavy supply chain businesses, finds it easier to establish a commercial closed loop by deploying To B-oriented AI technologies.

In the AI Era, Liu Qiangdong Cannot Let Go of 700,000 'Brothers'

For Liu Qiangdong, the challenges of the AI era include not only leading the company through strategic transformation but also placement (redeploying) JD's hundreds of thousands of frontline employees amid technological changes.

To reduce costs and increase efficiency, JD is actively investing in unmanned delivery technologies. In November 2025, Liu Qiangdong revealed, 'JD will establish the world's first fully unmanned delivery station next April.'

Official data shows that JD's warehouses have already achieved automation, with autonomous trucks having been tested internally for over 700,000 kilometers and unmanned vehicles for last-mile delivery having been tested for millions of kilometers.

As unmanned delivery technologies gradually mature, JD Logistics will no longer require a vast number of frontline couriers.

In response, Liu Qiangdong stated, 'In the future, robots will handle all deliveries. There will be no need for human couriers; it will definitely be robots doing the deliveries. However, I cannot let our 700,000 brothers go hungry or without work.' To prevent the unemployment of hundreds of thousands of couriers, Liu Qiangdong is leading JD in deploying 'safety net' businesses.

At the APEC CEO Summit China Forum held in June 2026, Liu Qiangdong said that JD has recently proposed the 'Nirvana Plan' internally, hoping to provide technical training to its existing 700,000 couriers and other blue-collar workers to help them cope with the impact of AI and technological advancements.

Previously, in an internal speech, Liu Qiangdong revealed that JD plans to establish over 80 robotics bases across the country to provide blue-collar workers with skills training in robot maintenance and repair. In the future, these blue-collar workers can provide maintenance services for users' robots.

In essence, self-operated e-commerce, To B-oriented AI businesses, and the 'Nirvana Plan' reflect Liu Qiangdong's consistent business philosophy—in a fiercely competitive market environment, building infrastructure that is difficult to replicate through asset-heavy investments and strong organizational capabilities.

While AI can reduce information processing costs and enhance decision-making and execution efficiency in the market, once it enters the real world, supply chains, operations, and maintenance, as well as end-point services, still require significant infrastructure and organizational capabilities to support.

Due to the high investment, long cycles, and slow returns of such businesses, many asset-light internet companies are unwilling to deeply invest in them. In contrast, JD has continuously invested in its supply chain over the years, providing rich application scenarios and practical experience for the implementation of asset-heavy AI technologies.

Therefore, from the perspective of the three-step AI strategy of 'digital intelligence, attached intelligence, and embodied intelligence,' JD is not merely betting on AI tools or model capabilities but attempting to drive the comprehensive intelligence of its accumulated supply chain, logistics network, and frontline service system. On the one hand, it aims to improve industrial operational efficiency by transforming warehousing, distribution, retail, and other links through AI; on the other hand, it seeks to transition hundreds of thousands of frontline employees to new roles in robot maintenance, intelligent services, and other fields through initiatives like the 'Nirvana Plan.'

In essence, JD is attempting to prove that competition in the AI era is not solely centered around model parameters and traffic entry points but also occurs in the infrastructure connecting the digital and physical worlds.

However, challenges remain. While its asset-heavy system is JD's advantage, it also brings significant uncertainty. If the core of future AI competition primarily focuses on large model capabilities and user entry points, JD may face the risk of marginalization.

However, if AI ultimately needs to penetrate real industries and complete the connection from the virtual to the physical world, then the logistics and supply chain system that JD has invested hundreds of billions of yuan in building over the years may no longer be just a cost burden but could become a key asset for the company to redefine its value in the AI era.

In other words, JD's true bet in its AI strategy is not whether it has large models but whether its heavy supply chain system can undergo a revaluation.