(Yicai) Sept. 7 -- Competition among Chinese office agent providers is shifting from desktop environments to the enterprise market, which offers stronger payment capabilities and higher entry barriers, with those that can make artificial intelligence understand and integrate into a company's business processes having the biggest potential to win.

Alibaba Group Holding, Tencent Holdings, Baidu, ByteDance, and other Chinese tech giants have been actively releasing new features and products for office agents since June, even adjusting their organizational structures to accommodate such changes. The battlefield is expanding towards large companies' customer service, ecosystem accumulation, and even computational infrastructure.

For example, Alibaba's QwenWork and Tencent's WorkBuddy have been integrated with their respective workplace messaging apps DingTalk and WeCom. In addition, ByteDance restructured its organization, incorporating the Lark, TRAE, and Coze teams into Doubao within a month and launched Doubao Work.

While office agents can improve the efficiency of individual roles, collaboration among people and between humans and agents may still constrain organizational workflows, Shu Junliang, vice president and head of product research at QwenWork, told Yicai. This has made upgrading office agents from personal tools to instruments that drive transformation in enterprise workflows and organizational structure essential, Shu added.

Rather than competing for entry traffic, leading tech firms are focusing on securing the resources agents need most once they enter an organization: enterprise context. The objects office agents need to understand are expanding from individual work habits and documents to various aspects of the organization, including organizational relationships, permissions, customer information, production, and financial data.

Enterprise context is a key area for future exploration in developing office agents, Shu pointed out. Transforming accumulated enterprise business data, workflows, organizational structures, and employee experience into information AI can understand is becoming increasingly critical, Shu stressed, noting that this involves enabling agents to automatically recognize employee identities, project backgrounds, and related materials while atomizing AI capabilities and embedding them seamlessly into existing systems.

QwenWork has been deeply integrated with DingTalk, while WorkBuddy has formed connections with WeCom, Tencent Meeting, and other traditional office services. ByteDance has strengthened the association between models and agents, while Alibaba plans to further connect QwenWork with AI models and computing resources from Alibaba Cloud, positioning them as foundational assets to win the competition.

The future office agents competition will not be limited to isolated comparisons of individual applications, but will evolve into a full-stack AI contest composed of models, computing power, cloud services, agents, collaborative software, and enterprise data.

"Office agents are merely a phase in product evolution, and the market may experience new changes every three to six months," Shu said. In the latest enterprise AI competition phase, the focus is no longer on a single product, but on the ability to continuously assist clients of various sizes in achieving AI transformation, Shu noted.

For example, a key step in the internal push for AI adoption at state-owned Chinese carmaker Changan Automobile is having its senior executives actively use AI, followed by key leaders in subordinate departments driving its implementation.

"At Changan Auto, agents like QwenWork are not just productivity tools for individual employees but also powerful instruments for building an AI-native organization," Liang Guangzhong, assistant president and general manager of the transformation and efficiency department, said to Yicai.

Companies are looking forward to the creative ways of working and productivity boost brought by AI.

AI agents can improve efficiency in certain work scenarios by several or even tens of times, noted Jing Xiaofei, deputy general manager of the transformation and efficiency department at Changan Auto. The carmaker aims to analyze input-output more systematically to ensure the benefits gained from increased efficiency and effectiveness exceed its investments, he added.

Monthly visits to 17 desktop AI office agents in China exceeded 60 million in June, up from 20 million in March, according to statistics from Analysys.

Editors: Tang Shihua, Martin Kadiev