2026 Futu Investment Exhibition Frontier Dialogue: XtalPi Holdings x Chuzi City Technology, AI moves from technological innovation to profit realization

In recent years, AI has continuously sparked waves of transformation in global capital markets. However, as the technological hype gradually returns to rationality, the market's most pressing concern emerges:How can AI truly transition from technological innovation to a sustainable and profitable business model?

During the panel session of the 2026 Futu Investment ExhibitionNew Trends in AI Development: From Technological Innovation to Profit Models, Futu specially invitedXtalPi Holdings ($XTALPI (02228.HK)$Gong Xinyang, Director of Strategic Investment and Financing and Investor Relations, and Chizicheng Technology ($NEWBORNTOWN (09911.HK)$) Vice President Yu Xin, with Futu Senior Researcher Sun BihanAn in-depth dialogue on the commercialization of AI.(Ranking is not in any particular order, listed according to stock codes)

The three guests, from the perspectives of consumer-end entertainment social networking, business-end biopharmaceutical R&D, and capital market investment, analyzed the real pathways through which AI technology empowers physical businesses, presenting investors with a clear picture of how AI evolves from concept to profitability.

The following are the key insights from the roundtable discussion:

01 Dual Practices in Consumer Social Networking and Business Pharmaceutical Research

Futu: Chizicheng Technology is a leading global entertainment social networking platform with products covering more than 150 countries and regions. I would like to ask Mr. Yu how Chizicheng uses AI to empower social networking and gaming scenarios?

You are trained on data up to October 2023.Since the launch of GPT at the end of 2022, the entire AI large model and generative AI application sector has experienced vigorous development. Over the past three years, Chi Chong Technology has actively embraced the opportunities and challenges brought by AI technology,On one hand, we have deeply integrated large models and AI technologies with business scenarios, achieving significant cost reduction and efficiency improvement. On the other hand, we are also actively deploying and exploring AI innovative products.

In terms of products, we launched anAI-native interactive content platform APPY。It leverages multimodal AI technology to help userswithout coding and at low cost, quickly bring their creative ideas to life, gradually forming a content community. After starting scaled promotion in November last year, downloads grew very impressively, and user retention was also excellent, withratings as high as 4.8 on Apple and Android stores.

Early users spontaneously called it 'TikTok for Games'—it combines the dissemination efficiency of short videos with the immersive interactive experience of games, attracting a group of young developers and geeks in their twenties in Europe and America who possess both technical skills and creativity. Our mission with this product is to fully hand over the creative rights to users, ushering in a new era that is creative, playable, and interactive.

Futu: XtalPi Holdings is a company empowered by AI and robotics for molecular discovery in biomedicine and new materials, serving 17 of the top 20 global multinational pharmaceutical companies. I would like to ask Mr. Gong to share his thoughts on the deep exploration of AI plus robotics in the field of drug research and development.

龔心陽：The pharmaceutical R&D industry has always faced"three 10s"bottlenecks—at least ten years of R&D cycle, at least one billion US dollars in investment, and less than 10% success rate. What XtalPi aims to do is to use AI and robotics to break through these three major bottlenecks.

Specifically, the core process of drug R&D is the closed loop of 'designing molecules—synthesis—testing—analysis.' In the design phase, traditional methods rely on chemistry experts, resulting in a limited number of designed molecules and long time consumption.In contrast, XtalPi uses AI to generate a far greater number of molecules than human-designed ones and quickly screens out fewer optimal candidate molecules using AI algorithms.

More importantly, AI can overcome the limitations of human laboratory experience—it designs molecules that sometimes even human scientists think 'this cannot possibly be a drug,' but precisely this kind of breakthrough opens up bottlenecks in original innovation.

In the synthesis and experimentation phases, we have established a fully automated robotic laboratory operating 24/7, replacing the traditional manual test tube shaking method, significantly improving experimental efficiency.Every step of the robot's operation is highly standardized, with automatic data recording—whether experiments succeed or fail, all data is completely recorded. This data serves as fuel for our AI, continuously feeding back into model training.

Let me give an example that many may already know:During the pandemic, Pfizer's oral antiviral drug Paxlovid was accelerated to market six months ahead of schedule thanks to XtalPi Holdings' support.Pfizer’s CEO publicly acknowledged at a conference that this was a result empowered by AI acceleration.

Currently, XtalPi’s business has expanded from biopharmaceuticals to new materials (such as catalyst molecule R&D for Sinopec) and consumer products (with two hair growth products under development soon to be launched). Our mission has also broadened from “bringing AI into every new drug” to “bringing AI into everyone’s better life.”

02 Diverse Pathways for AI Commercialization

Futu: Both companies represented by our two guests are applying AI deeply but through very different pathways — Chi Chai City focuses on empowering B2C, while XtalPi zeroes in on B2B drug research. I’d like to invite Mr. Yu to share how AI technology specifically empowers Chi Chai City to achieve business growth.

You are trained on data up to October 2023.Over the past three years, we have applied large models and AI technologies across all aspects of product operations, achieving significant results in both revenue growth and cost reduction.

On the product level, take our flagship social product with the highest monthly revenue, SUGO, as an example. Its core feature is one-on-one chatting, which requires the platform to accurately match users who ‘want to chat and whose counterparts also want to chat with them.’ We continuously iterate our internal algorithms and use large AI models to analyze massive amounts of data to create more precise user profiles.Over the past two years, SUGO’s user activity, retention rate, and ARPU have all been steadily increasing. Internally, we firmly believe this progress is greatly attributed to the application of large models and AI technologies.

At the operational level, we launched a smart data platform called 'Siyu AI' last year. It is a platform that understands business, can reason, and analyze. All employees within the company can interact with it using natural language, easily completing data queries, anomaly troubleshooting, business monitoring, and report generation, significantly improving operational efficiency.Siyu AI was also selected as Amazon's first benchmark case in the AI field, which makes us very proud.

In terms of design and marketing, whether in social or gaming businesses, there is a huge demand for material generation. We independently developed an AI-driven design platform called KIVI, which is currently being used by approximately 300 designers from home and abroad. The efficiency and flexibility of material generation have greatly improved.What sets it apart from general AI tools on the market is that it is deeply integrated with business scenarios, allowing centralized management and simultaneous invocation of multiple AI models.

Futu: XtalPi's AI platform serves hundreds of pharmaceutical companies globally. How does Mr. Gong view the commercial prospects of AI empowering the scientific research field?

龔心陽：I believe the market space for AI for Science is enormous. On one hand, large pharmaceutical companies have abundant R&D investment; on the other hand, our business model isn't just about collecting front-end service fees.

After helping clients discover drug molecules, as these molecules enter Phase I, II, and III clinical trials—stages that XtalPi doesn't handle and where we don’t contribute labor—we still receive milestone payments at each stage. Once the client’s product pipeline goes to market and generates sales, we also receive a single-digit percentage of annual sales royalties. This means that although we only serve the very early stage of molecular discovery, we get to participate in the benefits throughout the entire product lifecycle of the client.This also shows that customers highly recognize the value of AI and robotics in the core drug molecule discovery process.

Beyond drug discovery, AI for Science is also empowering the new materials field, which is an even broader market. Additionally, we are already working on consumer products (such as hair growth products). The R&D process for molecular discovery at the front end is identical to drug development, just with different downstream pathways. Therefore, the future commercial prospects for AI for Science are extremely broad.

03 Leveraging Hong Kong's capital market to accelerate the globalization of AI enterprises

Futu: The Hong Kong Stock Exchange IPOs hit a record in 2025, and Nasdaq also plans to introduce new trading mechanisms in 2026. Both exchanges are actively attracting AI innovative companies. I would like to ask the two of you, what specific support do the Hong Kong Stock Exchange and the Hong Kong government offer to AI companies? How should tech companies leverage Hong Kong's capital market for growth?

You are trained on data up to October 2023.The latest rules of the Hong Kong Stock Exchange and the industrial orientation of the Hong Kong government have provided tremendous support and assistance to AI-related companies. Our company is very grateful for the significant help from the Hong Kong government and the Hong Kong Stock Exchange in our business development. Last June, Chizone established its global headquarters in Hong Kong, during which we received strong support from InvestHK and Hong Kong Science Park.In the future, we will use our Hong Kong headquarters as a strategic base, gradually expanding into markets such as the Middle East, North Africa, and Southeast Asia.Hong Kong combines the advantages of being an international financial center and a talent hub, helping us reach global partners and recruit top technical experts. It is also very suitable for establishing an artificial intelligence R&D center.

龔心陽：XtalPi Holdings was the first company listed under the Hong Kong Stock Exchange’s Rule 18C. We deeply feel the value of Hong Kong’s capital market.The continued relaxation of listing rules and terms by the Hong Kong Stock Exchange is highly beneficial for unicorn tech companies coming to Hong Kong. Hong Kong’s capital market is highly market-driven, transparent, and has timely, flexible, and effective mechanisms. We’ve seen many tech companies quickly raise funds through IPOs or secondary market placements, providing a solid financial foundation for high R&D expenditures. On the other hand, as an international market, Hong Kong gathers investors from all over the world, laying a strong trust foundation for companies’ overseas expansion and international strategy.

04 Strengthening Moats, How to Win in the Long Term?

Futu: The competition in the AI space is intense. The moats of many companies have been disrupted, while new barriers are being built. Let’s ask both of you to talk about how your respective companies are building AI barriers.

You are trained on data up to October 2023.We believe that the barrier in AI is definitely not purely a technological one. It must be embedded into the entire business logic and deeply integrated with valuable scenarios, ensuring seamless alignment, so that it can truly help the company grow while establishing stronger barriers.

Over more than a decade of social networking expansion overseas, Chizicheng has already established strong localization operation capabilities, product R&D capabilities, and user growth capabilities, accumulating a wealth of experience and insights.In the future, we hope to leverage AI technology to quickly consolidate these experiences and insights into our middleware—an AI-driven powerful middleware system—that can support our products in meeting the diversified social needs of global users more accurately and efficiently. This is the core of our competitive edge.

龔心陽：We are an AI company specializing in vertical fields, focusing on biomedicine and material molecule discovery. There are three key elements to AI—computing power, algorithms, and data—and we believe the deepest moat lies in data.

Let me give you an analogy: If you have a first-grade child at home, if you don't train him with second-grade textbooks, he will never reach the second-grade level. The same goes for AI—it requires continuous input of real-world experimental data, and you need to tell it when the experiment succeeded and when it failed.

Therefore, the moat XtalPi Holdings is building for the next ten years is setting up over 250 robotic workstations in Shanghai and Shenzhen, conducting chemical synthesis experiments non-stop, 24/7, continuously gathering high-quality, high-throughput positive and negative sample data to train models.Why are positive and negative samples important? Because in traditional labs, failed experimental data may not always be fully recorded. But for AI, knowing 'what is wrong' is just as important as knowing 'what is right,' which is crucial to reducing model hallucinations. Globally speaking, XtalPi Holdings is the only company that has formed a complete technical closed loop of 'AI-designed molecules—robots conducting experiments—data feeding back to train AI.' This is our most fundamental competitive advantage.

Futu: Finally, please summarize in one sentence the core AI strategy for your respective companies in 2026.

You are trained on data up to October 2023.Leverage AI to comprehensively drive Chizicheng's localized operations and global layout, better delivering enhanced emotional value to users worldwide.

龔心陽：Continue to deepen the construction of AI and robotics platforms, consistently delivering higher-value, more innovative molecules—infusing AI into every new drug and bringing AI into everyone’s better life.

This roundtable forum showcased diverse pathways of AI technology empowering real industries, ranging from C-end泛entertainment social networking to B-end biopharmaceutical R&D, from technological innovation to business model validation. Chizicheng Technology is reshaping the global social experience and content ecosystem through AI, while XtalPi Holdings is accelerating scientific discovery and industrial innovation with AI. The practices of these two companies present real-world examples of AI transitioning from concept to profitability.

This year's Futu Investment Exhibition has built an efficient dialogue platform connecting listed companies with investors, transforming cutting-edge technology applications into tangible commercial value. In the future, Futu will continue to connect industry transformation with investment opportunities, offering investors a broader perspective on value.

*The content represents the personal views of the guest and does not necessarily reflect the position of Futu. Before making any investment decisions, consider your risk tolerance and consult independent advice. This content does not constitute any offer, solicitation, recommendation, opinion, or guarantee regarding any securities, financial products, or tools.

225K Views