After missing out on the AI era, will JD.com no longer play a minor role in embodied intelligence?

"Buy robots on JD.com."

At the World Robot Conference in August 2025, JD.com put up a huge poster, which also labeled its position in the current hottest embodied intelligence track - "Robot trading platform".

In the embodied intelligence track, JD.com doesn't make robots itself, but focuses on two aspects: trading and data.

A few days ago, JD.com also held a high - profile press conference on the embodied intelligence ecosystem, announcing the world's first embodied intelligence data infrastructure covering the entire chain of "acquisition, storage, labeling, training, evaluation, simulation, and testing". It also claimed that it is making every effort to promote the transition of embodied intelligence from laboratory research and development to large - scale commercial implementation, creating the so - called "super supply chain for embodied intelligence".

JD.com doesn't lock its target on the technical research of embodied intelligence or large - scale investment layout. Instead, it chooses to focus on trading. In contrast, leading Internet companies have established embodied intelligence laboratories, invested in leading enterprises, or got involved themselves.

This inevitably makes people feel that JD.com may miss out on the technological competition again.

Since the AI war began in 2020, JD.com has rarely released AI products among leading Internet companies. Whether it's consumer - facing large models or technological investment, JD.com is in the "others" category.

On the other hand, with the popularization of AI technology, leading cloud service providers are also competing secretly. As a latecomer, JD Cloud has never broken out of its small circle, with a negligible market share and is also listed in the "others" category.

This has led to a situation where JD.com is left behind in the entire chain from cloud computing to AI and then to robots.

Is there no "robot" in JD.com's embodied intelligence track?

Since last year, JD.com has set its sights on the embodied intelligence track.

In late March 2025, Liu Qiangdong, the founder of JD.com, visited several artificial intelligence projects at the Hong Kong University of Science and Technology. In the following months, JD.com carried out a fierce but short - term investment layout in robots.

In July, JD.com invested in three embodied intelligence companies in one day.

On July 21, JD.com announced investments in three embodied intelligence companies: Qianxun Intelligence, Zhujidongli, and Zhongqing Robot. In addition, JD.com also invested in Zhiyuan Robot in May last year. On July 30, JD.com led the investment in RoboScience, an enterprise engaged in the research and development of embodied intelligence technology.

Actually, JD.com has not invested in many robot enterprises. Besides the five mentioned above, there are also Pacini, Zhongke Guiji, and Yunshenchu. And the acquisition of Guo Qing's Xianglu Robot in 2024 represents all of JD.com's investments in robot enterprises.

It's worth noting that we previously observed that Meituan has carried out extensive layout in embodied intelligence in the past few years, including leading companies such as Unitree Technology, Zibianliang Robot, Yinhe Tongyong, Flexiv, and Tashan Zhihang.

Meituan has almost bought leading robot companies through investment and entered at an early stage with a large - scale shareholding.

When JD.com suddenly invested in three robot companies in July last year, JD.com and Meituan were in fierce competition in the food delivery business. It's unclear whether JD.com's sudden focus on embodied intelligence was a strategic detour or a new - field layout inspired by Meituan.

At the press conference on the embodied intelligence ecosystem, JD.com also released its self - developed acquisition terminal JoyEgoCam, the embodied large model JoyAI - AI, and the embodied intelligence data trading platform.

And a month ago, JD.com announced the completion of the world's largest and most comprehensive embodied intelligence data acquisition center, accumulating over 10 million hours of high - quality data in two years to help the embodied intelligence industry move from algorithm simulation to a new stage driven by real data.

With the release of the self - developed acquisition terminal JoyEgoCam this time, JD.com's embodied intelligence layout can be divided into two actions: data set acquisition and "buy robots on JD.com".

In contrast, companies involved in hardware products such as Xiaomi, XPeng, and Honor have all released products related to embodied intelligence.

In addition, Alibaba has been investing in the robot field for many years. Last year, Alibaba's management issued a statement saying that a group on embodied intelligence was established within Tongyi Qianwen. In February this year, Alibaba's DAMO Academy officially released the basic model of the embodied intelligence brain, RynnBrain. Recently, Gaode also launched the embodied intelligence hardware entity, the "Tutu" robotic dog.

In addition, ByteDance is also taking multiple approaches, including self - developing robots, investment layout, and algorithm support.

Missing out on the AI era

We believe that the technological competition from cloud computing to AI and then to robots is almost in line for leading enterprises. The advantages in the cloud era will continue to AI computing power, and the advantages in AI computing power and algorithms will continue to robots. On the contrary, if a company fails to perform well in the cloud era, it will be restricted in the subsequent stages.

JD.com's weakness in the cloud era has led to it missing out on the AI era, leaving it with limited ecological niches in the current embodied intelligence era.

In 2017, Liu Qiangdong shouted the slogan "Technology! Technology! Technology!" and vowed to lead JD.com to transform into a technology - based company comprehensively. He also said that JD.com would increase investment in fields such as graph networks, image recognition, big data, machine learning, and artificial intelligence.

This year was also the beginning of the previous wave of AI. The well - known event was AlphaGo defeating Lee Sedol in 2016. All enterprises were undergoing technological transformation and betting on AI.

However, after that, JD.com's investment in technology was interrupted due to the Minnesota incident in 2018 and the burst of the previous AI bubble.

For a long time after that, Liu Qiangdong was "out of position" and not leading JD.com at the front line. JD Cloud also made no progress in its listing for various reasons.

On the other hand, after Zhang Chen, the last CTO of JD.com, left in 2019, the position remained vacant, which also to some extent created a "generational gap" between JD.com and leading companies in terms of technology.

In recent years, there have been repeated reports of Liu Qiangdong's return. The first time was at the end of 2022 when Liu Qiangdong returned and competed with Pinduoduo on low prices in the e - commerce business. The second time was in 2025 when Liu Qiangdong frequently appeared in public and initiated a food delivery war. He also stated that JD.com would launch five new projects every year.

But we can clearly see that from 2020 to the end of 2022, several Internet companies entered the competition of large models. Since Liu Qiangdong had not returned to the front - line management of JD.com at that time, JD.com didn't even have a decent product to participate.

It wasn't until July 2023 that JD.com released the JD Yanxi large model. Different from the common large models suitable for consumer - facing products, "Yanxi" was originally JD.com's customer service product, which used intelligent customer service to solve the communication between consumers, merchants, and the platform. JD.com's Yanxi large model focuses more on industrialization.

Last year, robots became a hot track in the technology field, and JD.com also promoted cooperation with robot companies.

At the World Artificial Intelligence Conference in 2025, JD.com released the embodied intelligence brand JoyInside, which mainly provides large - model - driven dialogue capabilities for robots, robotic dogs, and AI toys.

He Xiaodong, the senior vice - president of JD Group and the deputy dean of JD Group's Exploration Research Institute, said, "We hope to build an open robot ecosystem to make the functions and experiences of robots better and sell more, achieving a win - win situation for consumers, robot brands, and JD.com."

A "technology company" without a CTO?

JD.com's lag in technology in recent years is also related to the long - term absence of a CTO.

JD.com established the CTO position in 2012, and it lasted for 7 years until Zhang Chen left in 2019. The CTOs were Wang Yaqing (January 2012 - December 2012), a 3 - year vacancy, Zhang Chen (March 2015 - March 2019). Earlier, Li Daxue was in charge of the front - end of JD's IT but did not actually hold the CTO position.

Currently, Cao Peng is the chairman of JD Group's Technology Committee, succeeding Zhou Bowen after he left JD.com in 2021. Zhou Bowen joined JD.com as the group's vice - president in 2017, in charge of the AI business. In 2019, JD.com established the Technology Committee, and Zhou Bowen served as the chairman.

Zhou Bowen joined at a very delicate time. JD Cloud was established in 2016, and in the following years, JD.com made large - scale investments in AI, IoT, and cloud services.

Before Zhou Bowen was Zhang Chen, who was the last person clearly in the CTO position at JD.com. By the time of Zhou Bowen, there was only the position of the chairman of the Technology Committee.

There is a difference between a CTO and a technology committee. In Zhou Bowen's words, a CTO is an in - house position leading the company's technological breakthroughs, while a technology committee is an ecological position, and one of its functions is to assist partners in solving technical problems.

Cao Peng's resume actually fits Zhou Bowen's positioning of the committee. It is a management position rather than a technical one.

Looking at the resumes of CTOs and core technology leaders of leading enterprises such as BAT, there are mainly two aspects. One is to make achievements in specific fields and be the technology leader or main contributor in the industry.

The second is to have a high status in the industry and be able to recruit relevant talents for the enterprise.

From this perspective, only Zhou Bowen has played such a role.

With JD.com missing out on the AI era, the market share of JD Cloud is not very impressive either.

JD Cloud operates in the dedicated cloud market, mainly serving industrial customers such as the financial industry. According to IDC data, in the first half of 2024, JD Cloud's market share in the Chinese dedicated cloud market was 10.3%. Based on the market scale of 17.49 billion yuan for dedicated cloud, JD Cloud's revenue in the first half of 2024 was 1.8 billion yuan.

Let's take Kingsoft Cloud as a reference. In the first half of 2024, Kingsoft Cloud's revenue was 3.67 billion yuan, and the full - year revenue in 2024 was 7.785 billion yuan. In addition, Kingsoft Cloud was listed on the Hong Kong Stock Exchange at the end of 2022. According to the data disclosed in its prospectus, Kingsoft Cloud's revenue in 2021 was 9.06 billion yuan, ranking 4th among Chinese cloud service providers with a 3.1% market share.

Crudely estimated, JD Cloud only has half of Kingsoft Cloud's market share, which is about 1.5% of domestic cloud service providers.

This article is from the WeChat official account "Big V Business", author: Zhou Ming. Republished by 36Kr with permission.