 |
HONG KONG, June 5, 2026 /PRNewswire/ -- On June 4, 2026, ARTi — DL Holdings' institutional-grade AI investment research platform — signed a subscription agreement with C Capital, a prominent Asia-Pacific technology investment firm. C Capital subscribed to 4.76% of ARTi's Class B shares (500 shares) at a price of USD 1.5 million. These shares carry no management or voting rights and entitle holders solely to financial returns. ARTi's pre-money valuation stands at USD 30 million. This Pre-Series A funding announcement marks the first time this AI application, quietly nurtured by DL Holdings, has come into public view — making a comprehensive introduction to investors both timely and necessary.
ARTi is an AI intelligent platform independently developed by DL Holdings, equipped with institutional-grade investment research capabilities. It has achieved structured, data-driven analysis with traceable investment reasoning, forming a complete closed loop of "judgment + recording + verification + capability accumulation" that meets the needs of investors of all types. ARTi currently provides full AI-agent investment advisory and AI-agent investment decision-making services, with AI-agent investment execution and other services to follow in the future.
DL Holdings began exploring AI as early as 2023, launching applications such as AI digital humans, AI smart advisors, and AI robots. Today, built upon a proprietary multi-agent AI technology architecture platform, ARTi is capable of delivering institutional-grade investment research. This represents a milestone breakthrough for DL in the AI domain. Leveraging vast, high-quality financial data and deep insight at the level of investment logic, ARTi has achieved institutional-grade investment research capabilities that far surpass the general-purpose large language models available on the market, enabling investors to pursue stable compounding growth.
I. What Sets ARTi Apart?
There is no shortage of general-purpose large language models on the market today. They are undeniably useful in high-error-tolerance contexts — translating documents, summarizing key points, getting up to speed on a new field. If prompted, they can even help analyze funds or stocks. But these models cannot be relied upon to guide real investment decisions. The reason is hallucination: general-domain error rates average around 5%, while in professional financial contexts that figure approaches 14%. ARTi systematically reduces the risk of AI hallucination in financial domains through multi-dimensional cross-validation of data and traceable reasoning logic.
How does DL achieve this? In short: DL's "3×8" framework — 8 AI analysts × 8 investment masters × 8-layer data architecture. This system's research capabilities already surpass those of general-purpose models available on the market.
1. The highest-quality data produces the most reliable results.
Inaccurate data renders analysis meaningless. Insufficient data inevitably creates blind spots. And without proper data organization, the hallucination problem in large language models cannot be solved. ARTi's real-time market data feeds, structured processing of listed company financial reports, global regulatory updates, and macro data all come from authoritative sources — the same data accessible to conventional vertical financial models. DL goes further, adding leading signals of corporate strategy, including recruitment data, patent filings, and government procurement records, as well as proprietary supply chain relationship data, equity structures, exchange announcements, and news. On top of all that, individual portfolio holdings, decision histories, and forecast accuracy records are incorporated. DL also brings over a decade of live trading data and decision logic accumulated through its family office practice, forming a powerful data moat.
2. The clearest logic produces the highest-quality algorithms.
ARTi's three-layer analytical logic is its core competitive advantage. The first layer consists of eight independently trained AI analysts covering macro, fundamental, technical, and risk assessment dimensions, alongside eight AI investment masters conducting independent analyses from their respective methodologies — a format dubbed "roundtable debate" — culminating in cross-validation and the surfacing of divergences.
The second layer evaluates the constraint conditions governing each investment conclusion. Every investment judgment must be grounded in cross-validated findings from the 8-layer data architecture. Each judgment must directly include the reasoning logic, underlying assumptions, and triggering factors — and every conclusion must be fully traceable, with no unexplained outputs.
The third layer records these verifiable conclusions. As each judgment is generated, the system automatically logs its content, timestamp, and market context, then continuously monitors it. As markets evolve, these judgments are either confirmed or refuted. In this way, ARTi's algorithms shift from "expressing opinions" to "verifiable objects." In other words, ARTi functions as the verification layer for investment judgments.
Compared with existing AI tools, this analytical system offers the following distinct advantages: outputs can be cross-validated to highlight key insights; the reasoning logic behind each judgment is recorded, and market developments are tracked against it to determine whether the judgment was right or wrong; and this history of judgments forms a highly personalized context for each investor — meaning the platform understands each user better over time.
II. DL Holdings at the Forefront — Again
DL's vision has never changed: Investing Made Simple. DL has always worked to share the service experience accumulated in the family office space with a broader base of investors. Whether through DL's digital family office, NeuralFin (突触科技), MiCang (米倉), AI digital humans, AI smart advisors, or AI robots — everything has been built toward that goal. In a word, DL has consistently chosen to do things that are hard but right.
Global financial institutions are all navigating the waves of AI, blockchain, and Web3.0. Robinhood has focused on the trading layer: in April 2023, Robinhood Connect bridged the gap between traditional centralized finance and the decentralized world; in October 2024 it launched prediction contracts tied to the U.S. presidential election; and in June 2025 it entered the blockchain securities space, enabling European users to trade over 200 U.S. stocks and ETFs on the Arbitrum network via blockchain-based tokens.
DL's internal discussions and instincts around decentralized applications — including prediction markets and tokenized financial assets — predate these trading platforms. But given Hong Kong SAR's regulatory realities and DL's mission to make investing accessible to everyone, DL chose the harder path: moving upstream from execution to decision support. DL wants every investor to have access to professional investment research capabilities and top-tier financial services — and that ambition is what gave rise to ARTi.
1. Institutional-grade research capabilities — where judgment compounds.
ARTi's value will not accumulate linearly; it will grow flywheel-style, and its moat will become extraordinarily wide. This is the market's reward for consistently doing what is hard but right. Just as Tencent's position in social networking is theoretically replicable — anyone could build a WeChat or QQ clone with AI today — no one can replicate the stickiness of WeChat's massive user base. ARTi is building the same kind of compounding dynamic: as users engage with ARTi, their judgments are recorded, outcomes are verified, verification results are analyzed to distill methodology, methodology comparison drives continuous system optimization, and system optimization produces better future judgments. This is the flywheel effect. Ultimately, ARTi's data moat, methodological moat, and network effects will become its most formidable competitive defenses.
2. ARTi's memory mechanism — making time itself a barrier.
A user who has spent three years with ARTi possesses three years of judgment records, verification data, and personalized feedback curated by ARTi. No competitor can replicate that in the short term — because the user's history of judgments is the most valuable investment asset they own, and ARTi has structured it all, turning it into a solid foundation for every future investment decision. Time itself becomes the most durable barrier.
ARTi currently covers three major markets — A-shares (Shanghai and Shenzhen), H-shares (Hong Kong SAR), and U.S. equities — with plans to expand into Japan and other Asia-Pacific markets, while also broadening its product coverage to include bonds, ETFs, and other mainstream financial instruments.
Looking ahead, ARTi will progressively enable investors to connect their own legitimate investment accounts, allowing them to leverage ARTi's research capabilities and judgment loop to execute trades autonomously via AI agents — including portfolio construction, automated trading strategies, and order placement. For retail users, ARTi can be integrated directly into Telegram, WhatsApp, and Web Apps; for professional users, CLI and API access will be available; for institutional partnerships, ARTi offers white-label solutions and a Skills API.
III. ARTi Empowers DL's Three Core Business Segments
1. Family Office
The family office segment serves ultra-high-net-worth individuals with investable assets of USD 30 million or above. ARTi's "AI Investment Committee" is already capable of assisting and participating in DL's family office investment decisions. In the future, ARTi will open access to family offices and investment institutions in Hong Kong SAR and globally, offering API integration services.
2. Professional Investors
DL Securities' core client base consists of Professional Investors in Hong Kong SAR — individuals with investable assets exceeding USD 1 million. Through serving them, DL has developed an acute understanding of their demand for high-quality investment decision-making, as well as the efforts they make to sharpen their investment capabilities. In the future, ARTi will deliver AI investment research services directly to all Professional Investors through DL Securities, with plans to extend the same quality of service to other domestic and international brokerages. This will meaningfully enhance brokers' ability to serve their clients — a win-win for all parties.
3. Retail and Prospective Investors
In serving retail and prospective investors, DL's NeuralFin (突触科技) is approaching the one-million-user milestone, and MiCang (米倉) — the app targeting mainland Chinese users — is now live on app stores. In the future, all of these users will be able to experience the full power of ARTi firsthand.
ARTi's Ultimate Vision: Infrastructure for the Investment Judgment Layer
In today's investment chain, the information aggregation layer has mature incumbents like Bloomberg — universally recognized as the infrastructure of financial information. At the trade execution layer, exchanges and electronic trading systems in every market have become standard infrastructure. But the investment judgment layer — arguably the most important — remains wide open. ARTi aspires to become the foundational infrastructure of this layer, and at the right time, DL intends to open-source ARTi. That is DL's choice, and its responsibility. （Related： Dong Bang Co Ltd to Showcase Veterinary Pharmaceuticals at Indo Livestock 2026 ｜ Latest ）
ARTi is headquartered in Tokyo, with R&D centers in Tokyo and Silicon Valley, and offices in Hong Kong SAR and Singapore.
