Prospect of New Stocks | Jin Zhiwei: Over three years of losses totaling 531 million yuan, the first in the AI digital employee market still finds it hard to "win effortlessly."

With the halo of being the "top player in the Chinese AI digital employee market," Jinzhihui has once again knocked on the door of the Hong Kong Stock Exchange. On August 21, Zhuhai Jinzhihui Artificial Intelligence Co., Ltd. (referred to as "Jinzhihui") submitted its listing application to the Hong Kong Stock Exchange for the second time, intending to list on the main board of Hong Kong. The last application was submitted on December 15, 2025, but it expired on June 15, 2026, after six months. Eight months later, Jinzhihui returns, with the title of "No. 1 in the Chinese AI digital employee market," yet the prospectus reveals another shocking figure: as of June 30, 2026, the company has deployed over 2 million AI digital employees and served more than 1,600 clients, covering six major state-owned banks and over 130 securities firms. However, during the same period, it has incurred losses amounting to 531 million yuan over three years, with only about 72 million yuan in cash remaining, and liabilities related to redeemable preferred shares as high as 2.126 billion yuan. While 2 million AI digital employees are "working," the company itself has suffered a loss of 531 million yuan over three yearswhy has Jinzhihui found itself in such a predicament? 2 million AI digital employees garnering clients still resulted in a loss of 531 million yuan over three years. For Jinzhihui, having millions of digital employees is a business card, and a thousand governmental and corporate clients form the base for transformation; however, a business card does not equate to revenue, and the base still needs monetization. Opening the prospectus, the most eye-catching labels for Jinzhihui focus on tangible results. As of the end of June 2026, the company has deployed over 2 million AI digital employees, serving over 270 banks and 130 securities firms. Most mainstream financial institutions in the country are on the client list, while also extending into government affairs, telecommunications, and high-end manufacturing, including over a hundred Fortune 500 companies. Receiving bulk orders in the highly regulated financial industry is already a significant industry threshold. The fact that Jinzhihui has been able to showcase many benchmark cases in the financial sector indicates that the company's platform has been tested under genuine business pressure: from the early K-RPA process automation tools to the K-APA intelligent automation platform, and now to the latest Ki-Agent enterprise-grade intelligent agent platform, the product line has evolved from "script recording automation" to "large model-driven, multi-agent collaboration," forming a complete technical stack covering process mining, low-code orchestration, large model invocation, and full lifecycle governance. However, GMTEight has observed that these impressive deployment figures have not smoothly translated into revenue and profit. From 2023 to 2025, Jinzhihui's operating revenue was 217 million yuan, 243 million yuan, and 256 million yuan, respectively. The compound growth rate over these two years was merely 8.7%. The revenue in the first half of 2026 was 65.13 million yuan, a year-on-year increase of about 42%this growth looks impressive but is based on the low baseline of 45.98 million yuan from the first half of 2025. In contrast, the domestic enterprise AI solution market has seen a compound growth rate of 37.1% during the same period. This paints an intriguing reality: despite being the leader in this niche market, Jinzhihui's revenue growth has lagged behind the overall trend. The root cause of the growth bottleneck primarily lies in the revenue structure. Currently, over 70% of the company's revenue comes from project-based delivery. Many projects involve "one deal, one discussion": deploying implementation teams to conduct requirements research, process configuration, scenario debugging, and on-site acceptance testing based on specific customer business processes. Each expansion into a large client necessitates matching corresponding products, implementation, and operation and maintenance personnel. The advantage of this model is that it makes securing benchmark orders easier, but its shortcomings are equally pronouncedrevenue expansion heavily relies on personnel growth, making it difficult for marginal costs to dilute quickly, leading to a trap where "more staff, more projects, higher revenue, and greater losses" ensue. The pressure on profitability is even more evident. From 2023 to 2025, Jinzhihui reported net losses of 63 million yuan, 122 million yuan, and 346 million yuan, accumulating losses of 531 million yuan over three years. As of the first half of 2026, it was still in the red, recording a loss of 109 million yuan. The losses arise primarily from three expenditure components: continuously increasing investments in large model and agent research and development; sales and marketing expenses aimed at multi-industry expansion; and the hefty personnel costs associated with project implementation and operations. In the first half of 2026, while the companys revenue experienced some acceleration, it still failed to turn a profit. Meanwhile, as Jinzhihui continues to incur losses, its cash flow is becoming increasingly strained. From 2023 to 2025, the company saw net cash outflow from operating activities of 56.67 million yuan, 26.49 million yuan, and 59.16 million yuan, respectively. Its ability to generate cash from its core business remains negative. As of June 30, 2026, the companys redeemable liabilities had reached approximately 2.126 billion yuan, with cash and cash equivalents totaling just over 71 million yuan, indicating clear financial stress. Furthermore, it is worth noting that Jinzhihuis redeemable liabilities of 2.126 billion yuan are tied to preferred share arrangements; if its listing falls short of expectations, this could trigger redemption clauses. From this, it is evident that having millions of digital employees merely serves as an entry ticket; commercial viability is the ultimate test, which means that the narrative of Jinzhihui, the leading AI digital employee company, still requires verification through operational performance. Competing in a trillion-yuan market, being industry leader does not guarantee an easy win. Of course, to understand Jinzhihui's value and risks, one must also step outside the singular company perspective and appreciate the opportunities and challenges faced in the entire digital employee space. In recent years, the market for enterprise-grade AI solutions in China has been rapidly growing, fueled by enterprises' technological upgrades, government digital initiatives, and the sustained development of large AI models. Currently, enterprise AI applications are gradually transitioning from isolated automation to comprehensive intelligence, with AI digital employees becoming core representatives. AI digital employees combine RPA, NLP, ASR, OCR, and large language model technologies to possess capabilities for cross-system operations, language understanding, knowledge reasoning, and task decision-making. According to the data disclosed in the prospectus, the scale of the Chinese AI solutions market is accelerating, with the market for enterprise-grade AI solutions expanding from 14.3 billion yuan in 2020 to 69.4 billion yuan in 2025, achieving a compound annual growth rate of 37.1%. With the acceleration of digital transformation in enterprises and smart government construction, it is expected to grow at a compound annual growth rate of 47.0% from 2026 to 2030, reaching 476.3 billion yuan by 2030. Positioned in this thriving market, Jinzhihui holds a dual leading position as a top player in the domestic AI digital employee sector. In terms of revenue, Jinzhihui maintained its position as the No. 1 player in Chinas AI digital employee solution market for three consecutive years from 2023 to 2025; when measuring by the number of mid to large enterprises served, it also remains at the top of the industry. Particularly in the financial services sector, the company has ranked first in revenue for three consecutive years, establishing a solid industry barrier. However, high market activity does not equate to certainty, and being the industry leader does not mean the landscape is solidified. Behind Jinzhihui's promising market prospects and leading market share lie prominent structural risks that continue to limit the company's growth ceiling. On one hand, the market structure is highly fragmented. Although Jinzhihui ranks first under specific statistical scopes, the overall market CR5 is low, with the market share of leading players remaining in the single digits. In this field, there are not only several native RPA vendors, but also major internet firms, traditional ERP suppliers, low-code platforms, and vertical industry large model companies continuously entering, and the fierce competition could reshape rankings at any time. On the other hand, there is currently no unified business model that has proven successful. Looking at the domestic and international landscapes, no enterprise-grade intelligent agent vendor has yet delivered a perfect profitability answer. Whether it's project-based, subscription-based, hybrid models, industry suites, or task-based billing, the entire industry is still in a phase of collective trial and error. In addition, the pace of technological iteration is rapid, presenting risks in direction. Generic large models, industry large models, lightweight small models, agent architectures, and multi-agent collaboration solutions are all evolving quickly. Todays investments in building platforms may face product depreciation and reduced competitiveness in a few years if they cannot keep pace with changes in technological paradigms. Continuous massive R&D investments serve as an entry ticket, but they do not guarantee that all research directions are correct. From all of the above, it is clear that while the potential rewards of the industry are tempting, the fragmented landscape, undefined business models, and technological iteration present crucial hurdles that Jinzhihui must overcome. In conclusion, Jinzhihui is a "growing company in a good sector"the industry outlook is broad, and its market position is leading, but there are significant shortcomings in financial health, business model efficiency, and governance structure. This also means that Jinzhihui's listing in Hong Kong is not the grand finale of this story; rather, it is the start of a new round of public examination. Its future investment value lies not only in the leadership it currently holds but in whether this company, with its first-mover advantage, can navigate through the fierce industry waves, overcome the mountain of ongoing losses, and achieve a transformation from benchmark cases to a standardized product platform.