應用商店
錢包

AI原生銀行:五大項目重塑金融行業

AI原生銀行:五大項目重塑金融行業

AI原生銀行——從零開始圍繞人工智能構建——正在自動化客戶服務、信貸、合規等多個領域。看看Catena Labs、One Zero、Bunq、微眾銀行以及中信aiBank等先行者是如何重新定義全球金融,並挑戰傳統銀行業務。

金融業的AI應用發展迅速。2010年代,不少銀行開始引入機器學習技術至信貸評分和聊天機械人應對客服,嘗試在現有框架下測試AI的潛力。至2020年,領先銀行已將進階算法應用於風險管理和客戶個人化。根據最新行業調查發現,65%的銀行計劃於2025年推出AI驅動的客戶服務——足證AI在銀行領域已成主流。然而,大部分這類努力仍只是在舊有系統上拼湊AI功能。相反,「AI原生」銀行希望打造以AI能力為核心的金融機構,徹底重塑銀行的營運模式。

AI原生銀行的概念正逐漸受到矚目,因為創業者和技術人員意識到即使數碼化的新型銀行,在應對AI為本世界時也有不少限制。傳統銀行的營運流程與基礎設施動輒幾十年歷史,經常被形容為「緩慢、昂貴、摩擦高、不靈活且難以配合」AI帶來的新機會。這促使初創公司及前瞻性的金融企業推動AI優先的銀行架構。

這些新銀行設計系統時,AI負責從客戶互動、欺詐監控,到信貸決策甚至合規管理,全部受人類監督。

什麼是AI原生銀行?

AI are conquering banks, Gorodenkoff/Shutterstock

簡單來說,AI原生銀行是自一開始就圍繞人工智能打造的金融機構,而非在傳統架構上強行加插AI。

有金融科技初創將AI原生銀行形容為「以AI為核心而非事後才加設AI的銀行」。

實際操作中,銀行所有產品、服務和內部流程都以AI演算法及自動化為運作主體,日常工作中只需最低程度人工干預。人員重點負責監督、戰略指導及處理特殊情況,但常規決策及互動由AI主導。

AI原生銀行的特色,是端到端全數碼化操作,AI管理客戶開戶、風險評估、交易及客戶服務。

先進機器學習模型分析客戶數據,實時提供個人化理財建議或偵測詐騙。聊天機械人和虛擬助理處理大量客戶查詢。更重要的是,這類銀行還會採用嶄新AI技術,例如用生成式AI打造對話介面,或以強化學習優化投資策略。目標是令銀行持續自我學習和適應,隨數據積累不斷改善服務——這是傳統舊系統難以做到的。

另一大特色是合規與風險管理直接嵌入AI系統設計。在傳統銀行,合規通常由獨立層級檢查及報告,甚至依賴手動操作。而AI優先銀行則自源頭起就以軟件程式確保合規,例如自動化可疑活動監控。「合規與法規風險理解必須與產品及工程同等重要」,Neville強調,這代表這類銀行直接將法規邏輯寫進AI工作流程中。

值得指出,「AI原生」並非等同「全AI化」。人類監督極為關鍵。

願景上並非要出現一家完全自動化、零僱員的銀行,而是極高自動化水平下,人與AI緊密協作。例如,有AI銀行項目計劃以「AI數碼員工」負責內部工作(如撰寫軟件),而人類負責監督及高階決策。對外服務時,AI助理會解答常規問題,若遇到需要同理心或判斷的狀況才轉交人員處理。

下文將介紹五個體現AI原生銀行理念的項目。

Catena Labs — 構築「AI經濟」專屬銀行

Catena Labs

討論度極高的新項目之一,是美國初創Catena Labs,由Sean Neville(因創辦Circle及USDC穩定幣而知名)共同創立。

Catena Labs於2025年5月宣布獲得1800萬美元種子輪融資,目標是構建一間「全受監管、AI原生的金融機構」,鎖定新興「AI經濟」發展。

此輪融資由Andreessen Horowitz旗下a16z crypto基金領投,參與方有Breyer Capital、Coinbase Ventures及NFL明星Tom Brady——這一資金及投資人組合足見此項目備受關注。

Catena願景相當雄心:要建立一間銀行,讓AI系統(統稱為「AI代理」)可擁有賬戶、自主執行交易,並能與其他代理或人類進行財務互動。Neville認為,不久將來「AI代理將主導大部分經濟交易」,但傳統銀行完全無法應對。

例如,交易算法或電商機械人可能需代表人類每秒完成數千次支付或簽署合約——這些對現有銀行流程而言極難應付。

Catena的解法是從零重建金融基建,以滿足這些需求。

Catena方案重心,是以穩定幣——尤其是Neville共同創立的USDC——作為交易的「AI原生貨幣」。

穩定幣運行於區塊鏈網絡,支援全球跨境近乎即時且可編程支付。Catena Labs主張,這特別適合24/7全球運作的AI代理,既要高速、低成本交易,亦要避免人為延誤。活用USDC等數碼貨幣,該銀行將讓AI客戶像傳輸數據一樣靈活轉賬,同時嚴格執行KYC及AML等合規要求。

Catena Labs對合規與信任尤為重視。

Neville強調,獲得正式銀行牌照及做到合規是整個項目重點。銀行將由「AI運作,人類監督」,即自動化管理日常流程,由人類設定政策和在需要時介入。Catena亦發布了Agent Commerce Kit(ACK)——一套開源工具包,可驗證及管理AI代理身份。為AI代理創建可信數碼身份,是金融合規最大難題之一,ACK嘗試提供協議以登記和認證AI代理於金融交易中的身份。

Catena Labs毫不諱言指現有銀行體系種種問題逼使新方案誕生。現行全球金融基礎建設被形容為「緩慢、昂貴、全球摩擦多、不靈活,難以回應AI帶來的新機遇與風險」。

Neville認為,傳統銀行還會主動阻礙自動化代理——不少系統本身為偵測及封鎖「機械人」而設,偏偏這讓合法AI代理難以參與金融活動。Catena構想的銀行剛好相反,目標是「AI代理作為主要用戶,而不是封鎖他們」。

截至2025年中,Catena Labs仍處於開發階段——未有公開產品,正積極申請牌照。1800萬美元資金將用於擴大團隊及產品研發。由於Neville有Circle背景,可見該創業項目很大機會會跟監管機構緊密合作(或尋求銀行牌照、或與現有銀行合作),確保AI原生銀行合法穩健地推向市場。

One Zero Bank — 以色列AI驅動數碼銀行

Ori Goshen, One Zero Bank

部分AI原生銀行項目剛起步,位於以色列的One Zero Bank已正式運作,並深度整合AI至其服務。

One Zero於2022年底開業,是以色列首間全數碼銀行——亦是逾45年來第一家獲頒銀行牌照的新銀行。

由著名科技人、Mobileye(自動駕駛領導企業)創辦人Amnon Shashua教授聯合創立。在雄厚資金支持下,One Zero Bank一開始就將AI技術與銀行業結合。銀行於開業時表示其模式「由人工智能驅動,融合傳統與新型銀行之優勢」。實際上,One Zero將數碼便利性與私人銀行式體驗結合,利用AI強化客戶服務與個人化。 capital](https://www.calcalistech.com/ctechnews/article/byi4zgrlkx),彰顯對其模式的信心。到2025年,該銀行已籌集約2.42億美元,估值約為3.2億美元,投資者包括像騰訊這樣的科技巨頭和來自軟銀生態系統的金融科技基金。

AI 是 One Zero 客戶體驗的核心。

2024年2月,該銀行推出了「Ella 2.0」,這是一個以生成式 AI 為基礎的服務平台,為客戶提供虛擬理財助理服務。這項服務是與 AI21 Labs(專注大型語言模型的以色列 AI 初創公司)合作開發而成,本質上就是一位全天候 24/7 的 AI 私人銀行家。

客戶可以用自然語言與 Ella 互動──無論是查詢跨帳戶的複雜財務問題、尋求預算建議,還是解決問題──都能即時獲得具情境感知的回應。系統支援多種語言,並以大量銀行相關問題進行訓練,提升準確度。

根據該銀行表示,Ella 2.0「能夠瞬間回應、全年無休運作,並利用機器學習為用戶度身訂造個人化的金融服務。」換句話說,系統會不斷從客戶互動中學習,提供更佳協助,同時有真人銀行專員隨時待命支援。

One Zero 首任行政總裁 Gal Bar Dea 強調這個 AI 助理大大提升了服務質素。他說:「Ella 2.0 的能力超越語言障礙,確保即時、準確及個人化的回覆,同時不斷進化,以切合每位客戶的需求。」

One Zero 引以為傲的,是帶領銀行業「由實驗性生成式 AI 跨越到實際應用」這場全球變革。

AI21 Labs 聯合行政總裁 Ori Goshen 表示:「One Zero 全新的 AI 助理 Ella 代表數碼銀行業走向更佳客戶體驗的轉變──變得更快、更可靠,也更貼合每位用戶。」

這些調子彰顯科技初創與銀行在 AI 方案開發上如何深度整合。

除了 Ella,One Zero 還在幕後以 AI 處理多方面運作。自動化算法處理大量銀行日常營運和決策。例如,AI 模型用於信貸風險評估及投資推薦,能從數據中學習並不斷優化結果。

該銀行策略是盡量自動化日常工作,從而降低成本,並能提供更有競爭力的收費。

同時,One Zero 仍然保留真人財務顧問給客戶諮詢(銀行承諾結合「個人理財經理」與 AI 助理的混合支援)。這種雙軌方式切合了既想善用 AI 效率、又希望有真人專家協助重大決策的客戶需求。

One Zero 在 AI 方面的大量投資,已經轉化為客戶參與度的成果。

有報導指出,AI 助理在推出後不久已能獨立處理高達40%的客戶查詢,並協助真人專員處理其他大量問題。這大大縮短了回應時間──銀行聲稱已消除大部分查詢的等候時間──確保客戶隨時獲得一致、高質素的答案。

AI 甚至能處理複雜的交互式查詢,例如:One Zero 曾舉例,有客戶問「我同朋友在倫敦去過的那間印度餐廳是哪一間?」,系統能推斷並找出相關交易。這種能力體現了結合交易數據與對話式 AI 的威力。

從市場角度來看,One Zero Bank 成為新銀行如何靠 AI 別樹一幟的典範。在以色列競爭激烈的銀行界,One Zero 的賣點不只限於擁有出色的手機應用──很多銀行都做得到──而是服務更加智能和主動。銀行會透過 AI 分享異常消費警報、預測現金流,或建議財務行動,這些全都是 AI 以用戶數據分析而得。這正好切合更廣泛的趨勢:消費者日益期望類似 Netflix 或 Spotify 客製化娛樂體驗的即時、個人化金融服務。One Zero 就是運用 AI,定位做「金融管家」。

One Zero 仍有挑戰,特別是計劃擴展以色列以外市場。該行原本有國際擴展計劃,但因 2023 年底地區局勢等外在因素而暫緩部分動作。

儘管如此,他們的發展仍受全球關注。如果 One Zero Bank 持續獲得成功,很可能會啟發更多國家出現以 AI 為核心的數碼銀行。這同時也為監管機構提供了一個現場示範,證明 AI 可以安全地融入銀行業運作。值得一提的是,以色列監管機構已向 One Zero 頒發全面銀行牌照,顯示對其經營模式及資本的信任──對其他希望獲批的 AI 原生銀行來說,是正面的信號。

Bunq – 歐洲首間 AI 驅動虛擬銀行

000006546746.jpg

在歐洲,早已成立並採用 AI 原生策略的代表之一就是 Bunq,這家荷蘭數碼銀行因科技主導、用戶為本的理念被稱為「The Free 之銀行」。

Bunq 成立於2012年,現時於歐洲各地擁有數百萬用戶。2023年底,Bunq 宣布已成為「歐洲首間 AI 驅動的銀行」,一時轟動業界。

Bunq 把生成式 AI 深度整合至平台,遠超同業,目標是徹底改變客戶與理財互動的方式。當中最重要的項目是「Finn」──Bunq 的 AI 個人理財助理。

2023年12月,Bunq 將 Finn 作為面向客戶的生成式 AI 工具整合進應用程式。

Finn 幾乎取代了 Bunq 應用內傳統的搜尋與導航功能。用戶無須再慢慢逐頁翻查菜單或交易清單,只需以自然語言向 Finn 提問或下達指令即可。「Finn 會令你大為驚喜,」Bunq 創辦人兼行政總裁 Ali Niknam 在發佈會上說,強調這是「多年 AI 創新」和「專注用戶需求」的成果。

正如 Niknam 所述,目標是「徹底改變你對銀行的認知」,令互動像日常對話般輕鬆。

Finn 到底有多強大?根據 Bunq 所述,非常多。用戶可以問「我上個月花了多少錢在買餸?」或「我每月平均水電煤開支是多少?」Finn 會即時剖析交易數據作答。它也能處理更複雜、跨多個資訊的查詢。

例如,Niknam 分享:「它甚至能整合數據回答超越單一交易的問題,例如『上星期六我在 Central Park 附近的咖啡店用了多少錢?』」AI 具備情境理解力,能明白「Central Park 附近的咖啡店」是指定商戶和日期的意思,是一般搜尋功能難以比擬的。這種對話式查詢方式,令用戶無需會計知識、亦不用再繁瑣地手動查找,便可以輕鬆分析自身消費和搜尋資料。

除了查詢功能,Finn 同時協助理財規劃與預算編制。用戶可以索取建議或見解,例如「我今個月可否有餘錢加500歐元入儲蓄?」並得到數據化回覆。簡直就像隨身的私人會計師。

Bunq 利用這點鼓勵客戶養成更健康的理財習慣。內部來看,Bunq 的 AI 亦會分析同一用戶多個綁定帳戶(利用歐洲開放銀行規範)的交易模式,提供全面財務總覽。這代表 Finn 不只可查到 Bunq 戶口的狀況,如用戶授權,更能連繫其他銀行,做到一站式財務管理──對預算與理財規劃來說極具價值。

Finn 的成效同樣顯著。

有報導指,Finn 能獨立處理約 40% 客戶查詢而無需人手介入,亦能協助處理另一部分查詢。

這大幅減輕了 Bunq 客戶服務團隊的工作壓力,也提升了用戶的回應效率。事實上,Bunq 指出於2024年初引入 Finn 後,用戶互動比以往更高效,許多查詢都由 AI 即時回覆。至於較複雜需真人跟進的問題,Bunq 團隊可專注處理,因為 AI 已預先篩走大部分簡單查詢。

這種模式,讓 Bunq 能隨用戶人數增長,仍然維持可擴展的客戶服務。

Bunq 擁抱 AI 之際,業績亦見迅速擴展──包括地區和產品層面。公司於 2023 年申請美國銀行牌照,準備進軍美國市場,而這類創新做法正好令 Bunq 在競爭日益激烈的新型虛擬銀行中脫穎而出。

同時值得留意,其他金融科技亦正開始追隨。例如美國新銀行 MoneyLion 於同期宣布新增 ChatGPT 搜尋功能,另有名為 Dave 的平台引入「DaveGPT」作客戶查詢。

不過,Bunq 率先將 AI 完全融入核心功能(甚至直接以 AI 取代搜尋),令其取態領先同儕。

從商業角度看,Bunq 利用 AI 不僅協助用戶,也用於發掘洞見,以推動新功能。通過分析用戶如何向 AI 查詢理財問題,Bunq 能掌握痛點或熱門請求,再針對性研發新功能或產品。

舉例說,若多位用戶詢問「到年底我夠不夠錢買X?」Bunq 便可能開發自動儲蓄規劃功能。這種依靠數據推動的創新,正是 AI 原生銀行的競爭優勢──從用戶互動到服務改進的反饋循環極之緊密。

但 Bunq 亦十分謹慎to couple AI with human oversight. All AI responses are monitored for accuracy and relevance.

結合人工智能同人手監察。所有AI嘅回應都會被監控,確保準確同相關。

The bank has emphasized that Finn’s advice is based on data but customers should exercise judgment – it’s an assistant, not a fully autonomous financial manager (at least not yet). Additionally, privacy and security are paramount; Bunq has to ensure that the AI only accesses data the user has permissioned and that sensitive information is protected. So far, no major issues have been reported, and customers have largely responded positively to the convenience of conversational banking.

銀行強調Finn嘅建議係根據數據,但客戶都要用自己嘅判斷——佢只係個助手,唔係完全自動化嘅理財經理(至少而家未係)。另外,私隱同安全係最重要嘅;Bunq要確保AI只可以接觸到用戶授權嘅數據,亦要保障敏感資料唔會外洩。到依家為止,未有重大的問題被報告,而大部分用戶都正面回應咁方便嘅對話式銀行服務。

Ali Niknam, Bunq’s CEO, has framed the AI push as part of Bunq’s mission to simplify banking. In his view, traditional banks burden customers with clunky interfaces and jargon, whereas Bunq wants to “make life so much easier” for users through technology.

Bunq嘅CEO Ali Niknam話,推動AI係Bunq簡化銀行業務使命嘅一部分。佢認為傳統銀行經常用啲難明介面同行業術語,令客戶覺得麻煩;但Bunq就希望利用科技「令用戶生活簡單好多」。

By making banking as easy as texting a friend, Bunq hopes to deepen customer loyalty and engagement. Indeed, industry analysis shows that personalization and ease of use significantly boost customer satisfaction in banking.

Bunq希望銀行服務可以好似同朋友send message咁簡單,從而提升客戶忠誠度同參與度。事實上,業內分析顯示,個人化同易用性可以大大提升客戶對銀行服務嘅滿意度。

Bunq’s AI strategy hits both targets: personalize the experience (since Finn’s answers are unique to your data and questions) and make it easy (no need to learn the app menus or finance terminology).

Bunq嘅AI策略做到兩個目標:一係個人化體驗(因為Finn每次回應都根據你嘅數據同提問獨一無二),二係簡單易用(唔使學app menu或者金融術語)。

As one of the first movers in AI-powered banking in Europe, Bunq offers a valuable example for the industry. It demonstrates that even an operational bank with millions of users can successfully infuse AI at the core of its services – it’s not just something for brand-new startups. Bunq’s experience will be closely watched by other European banks and fintechs. In a way, Bunq is turning into a tech company as much as a bank, continually integrating the latest AI developments. If Finn and subsequent AI features continue to perform well, it’s likely we’ll see more banks launching their own GPT-style assistants or AI-driven personalization features in an arms race to attract digitally savvy customers.

作為歐洲最早採用AI銀行業務之一,Bunq為行業帶來好好嘅示範。佢證明咗就算係有數百萬用戶嘅運營銀行,都可以成功將AI融入核心服務——唔只係初創公司先做到。Bunq嘅經驗預計會畀其他歐洲銀行同fintech公司密切關注。從某個角度講,Bunq都愈來愈似科技公司,唔止得銀行,持續引入最新AI進展。如果Finn同未來AI功能表現都好,相信好快會有更多銀行推出自己嘅GPT助手,或者利用AI推動個人化功能,掀起吸引數碼世代客戶嘅競爭。

WeBank – China’s Pioneering AI-First Bank

(略過圖片)

No discussion of AI in banking would be complete without WeBank, China’s trailblazing digital bank that has been a pioneer in AI adoption since its inception.

講到銀行業AI,點都唔可以唔提微眾銀行(WeBank),呢間中國先驅數碼銀行,由創立開始就已經走在AI應用最前線。

WeBank was founded in 2014 as China’s first internet-only bank, backed by tech giant Tencent. From the beginning, WeBank’s strategy was to leverage cutting-edge technologies – encapsulated in its “ABCD” mantra (AI, Blockchain, Cloud, Data) – to serve millions of customers at low cost. Over the past decade, WeBank has grown explosively, providing loans, payments, and financial services to tens of millions of users, many of them underbanked individuals and small businesses. Its success is often credited to its deep integration of AI in operations, enabling it to manage volume and risk far more efficiently than traditional banks.

微眾銀行於2014年成立,係中國首間純網上銀行,由騰訊做後盾。開頭已經講明要利用最前沿科技——即佢哋所講嘅「ABCD」(人工智能、區塊鏈、雲計算、數據)理念——以低成本為上千萬客戶提供服務。十年來,微眾銀行增長爆炸性,提供貸款、支付、金融服務畀數千萬用戶,其中好多都係冇銀行服務嘅個人同細企。佢哋成功地方,好多時都係因為AI深度融入日常運作,令處理大量交易同風險管理遠比傳統銀行有效率。

One of WeBank’s notable achievements is the extent to which it uses AI and automation in customer service and support. As of a few years ago, WeBank reported that it was receiving around 100,000 customer service queries per day, and its AI “virtual robots” were handling 98% of them without human intervention.

其中一個最明顯嘅成就就係佢幾咁大範圍用AI同自動化去處理客服支援。幾年前,微眾銀行已經話每日收到大約10萬宗客戶查詢,而佢AI「虛擬機械人」已經可以自動處理98%查詢,唔使人手介入。

These virtual agents use natural language processing and speech recognition – essentially early versions of the kind of AI that powers today’s voice assistants – to resolve customer inquiries. Dr. Yang Qiang, a chief AI consultant at WeBank, explained that they deploy facial recognition, voice recognition, and NLP to improve service and convenience. Customers can interact through chat or voice, and the AI can authenticate them (via facial recognition) and address issues or execute requests in real time.

呢啲虛擬客服大使會用自然語言處理(NLP)同語音識別——其實就係現時語音助手AI早期版本——去解答客戶問題。微眾銀行AI顧問楊強博士表示,佢哋用咗面容識別、語音識別同NLP,提升服務便利性。客戶可以打字或者語音同AI互動,系統亦可以即時用面容識別驗證身分,解決問題或者即時幫佢辦手續。

WeBank’s philosophy has been that AI is there to “augment, not replace” human service – a stance that sounds similar to Western banks, but WeBank has taken it to an extreme degree of implementation. “Automated service is not an enemy to human services. They should work side by side,” Yang Qiang told CNBC. The result is a highly scalable model: a relatively small team of human staff can oversee a customer base of millions because AI is doing the heavy lifting day-to-day. In fact, WeBank famously started with only a few dozen employees and no physical branches, yet it was able to disburse enormous volumes of micro-loans across China by relying on AI-driven credit algorithms and customer interactions through smartphones. This operational efficiency is a major reason WeBank turned profitable within just a couple of years of launch, a rare feat for a new bank.

微眾銀行理念係AI「增強而唔係取代」人類服務——口號好似外國銀行,但佢哋真係做到極致。楊強曾向CNBC表示:「自動化服務絕對唔係同人手服務做對立,兩者可以並存。」結果形成一個極具擴展性嘅模式:用細小人手監督幾百萬客戶,因為AI每日已經做晒大部分搵食工作。其實微眾銀行出名就係開頭得幾十個員工,冇實體網點,但靠AI信貸演算法同手機互動,可以喺全國發放大量微額貸款。咁高效率都係微眾銀行開業短短數年內已經賺錢(對新銀行嚟講極罕有)嘅主因之一。

Another area where WeBank shines is AI-driven credit risk analysis and loan approval.

微眾銀行引以為傲嘅仲有AI主導嘅信貸風險分析同批核。

Traditional banks often require lengthy paperwork and human underwriting for loans, but WeBank automated much of that using machine learning models. By analyzing vast amounts of alternative data – such as social media behavior, mobile payment history (leveraging Tencent’s ecosystem), and other digital footprints – WeBank’s AI can assess creditworthiness quickly and extend small loans to individuals and SMEs that might be rejected by larger banks.

傳統銀行做貸款好多時需要繁複文件同人手審批,但微眾就用機器學習模型自動化晒大部分流程。AI會分析大量新型數據——好似社交媒體習慣、手機支付紀錄(利用騰訊生態圈)、同其他數碼足跡——快速判斷信貸評級,批出貸款畀可能被大型銀行拒絕嘅個人或中小企。

This inclusive approach has extended credit to segments previously deemed too risky or costly to serve. Yang Qiang noted that such technology creates “the possibility for WeBank to have more efficiency than traditional banks in processing loans and conducting risk analysis”, which indeed has been borne out. WeBank can process loan applications in minutes and monitor them continuously, something legacy banks find hard to match.

咁包容嘅做法令之前被視為高風險或高成本族群都獲得信貸。楊強提過,呢類技術令「微眾銀行無論批核貸款或者分析風險都可以比傳統銀行更有效率」,而事實證明真係做到。微眾銀行可以幾分鐘內處理貸款申請,仲可以全天候監控,傳統銀行難以跟上。

WeBank has also been an innovator in AI research.

微眾銀行喺AI科研方面都係創新者。

It has invested in areas like federated learning, a technique to train AI models on sensitive data from multiple sources without compromising privacy. This was important for WeBank to collaborate with other institutions (like sharing fraud data) while respecting China’s strict data privacy rules.

佢投入咗好多資源發展「聯邦學習」(federated learning),即係可以保護私隱下,喺多個數據來源上訓練AI模型。對微眾銀行嚟講,好重要可以咁配合其他機構(例如共用詐騙資料),而且又唔會違反中國嚴格嘅數據私隱法例。

The bank’s technologists have published papers and open-sourced tools, indicating that WeBank sees itself as a tech leader, not just a financial services company. In March 2025, WeBank even shared a vision for an “AI-native bank” at a global conference, highlighting how a decade of its tech expertise is pushing banking to be “smarter and more inclusive.”

佢啲技術人員發表過多篇論文同開源工具,顯示微眾銀行已經唔只當自己係金融公司,更係科技行業領頭羊。到2025年3月,微眾銀行仲喺國際會議提出「AI原生銀行」嘅理念,強調十年科技經驗點樣推動銀行變得更聰明同包容。

This suggests WeBank is aiming to stay at the forefront of AI in finance, possibly exploring next-gen AI like generative models for even more advanced services.

可見微眾銀行目標係保持金融AI最前線,未來可能發展更先進、自動化更高嘅AI生成模型等新服務。

Despite its tremendous automation, WeBank hasn’t eliminated the human element. Instead, it has reallocated it. With AI doing routine work, human employees focus on areas like improving algorithms, handling exceptional cases, and developing new products.

雖然自動化程度極高,微眾銀行仍然保留人手參與。不過工作重點由日常事務轉去改良演算法、處理特別個案同開發新產品。

WeBank’s staffing strategy reportedly has about 60% of employees in technology roles – an unusually high ratio for a bank, but logical for what is essentially a fintech institution. This tech-first culture further cements WeBank’s status as an AI-native bank avant la lettre.

據報微眾銀行有成六成員工係做科技崗位——對一般銀行嚟講好罕有,但對純科技金融(fintech)行業就合情合理。呢種以科技為本的文化,亦鞏固咗微眾銀行作為「AI原生銀行」嘅地位。

CITIC aiBank – A Joint Venture of Finance and Tech

(略過圖片)

Around the same time WeBank was taking off, another notable experiment in AI-centric banking was underway in China: CITIC aiBank (often just called “AiBank”).

當微眾銀行開始冒起時,中國另一個以AI為核心嘅銀行新嘗試都誕生咗:中信百信銀行(CITIC aiBank,簡稱AiBank)。

This is a joint venture between China Citic Bank, a mid-tier commercial bank, and Baidu, the internet search and AI giant. Launched in late 2017, aiBank was established as a direct, branchless bank with the explicit goal of leveraging big data and artificial intelligence to deliver smarter financial services.

呢間銀行由中信銀行(中國中型商業銀行)同百度(中國互聯網搜尋同AI巨頭)合資創立。2017年底推出,明確以大數據同人工智能推動智能金融服務,屬於純線上、冇分行嘅直營銀行。

With a registered capital of 2 billion yuan (about $300 million at the time) and ownership split 70/30 between Citic Bank and Baidu, aiBank represented a blend of banking domain knowledge and cutting-edge tech capability.

AiBank註冊資本20億元人民幣(當時約3億美元),中信銀行佔七成,百度佔三成,集合傳統銀行經驗同最先進科技能力。

AiBank’s focus from the start was on lending to consumers and small businesses, segments often underserved by traditional banks in China. By using Baidu’s AI technology, aiBank aimed to develop new risk assessment models that could better evaluate borrowers who lack extensive credit histories. “AiBank will focus on lending to individuals and small businesses while leveraging big data and artificial intelligence to build new risk control models,” said Li Rudong, the bank’s president, at its launch.

一開始,AiBank重點都係針對消費者同中小企嗰啲傳統銀行唔太願意服務嘅族群。用百度AI技術,開發適合信貸歷史唔夠完善用戶嘅新風控模型。AiBank總裁李如東喺開業時講:「AiBank會專注為個人、中小企業提供貸款,同時用大數據、人工智能建立新嘅風險控制模式。」

This indicates that aiBank intended to analyze non-traditional data – possibly including search data, social data, etc., thanks to Baidu – to make credit decisions. The expectation was that AI-driven insights could identify creditworthy customers that legacy scoring methods might overlook, thus profitably expanding financial inclusion.

即係話,AiBank希望利用百度嘅優勢,分析非傳統數據(例如搜尋數據、社交數據之類),去做信貸決策。期望靠AI洞察力,搵到傳統評分法忽略咗其實信貸質素唔錯嘅用戶,擴大可盈利嘅金融包容性。

A striking detail revealed at launch was that 60% of aiBank’s employees would be tech staff. This was essentially unheard of in banking at that time and signaled how differently aiBank would operate compared to a typical bank where most staff are in branches or general operations. By concentrating on engineering and data science talent, aiBank put itself on a path to continuously develop and refine AI systems in-house. Baidu’s contribution was not just capital but also technology – including its AI platforms, cloud services, and perhaps even its vast user data (within privacy/legal limits). This partnership was part of a broader trend in China of tech companies and banks teaming up – similarly, Alibaba with MYbank, and Tencent with WeBank – to create hybrid entities that marry the strengths of each. In Baidu’s case, aiBank also offered a way to monetize its AI research in finance and showcase its AI leadership.

開業時有一個好誇張但重要嘅數字:AiBank 60%員工都係科技人才。呢個比例喺銀行界好少見,反映AiBank營運模式完全唔同於傳統銀行(大部分員工係分行或一般職能)。以工程、數據人才為主,等AiBank持續有能力內部開發升級AI系統。百度唔單只出資,仲帶來AI平台、雲服務、甚至可能帶動龐大數據(當然要守規例同私隱)。呢啲合作正係中國銀行同科技公司新趨勢——好似阿里巴巴同網商銀行、騰訊同微眾一樣——大家結合優勢,開創全新銀行物種。對百度嚟講,AiBank都係金融變現AI科研、展示AI領導地位嘅一個方法。

At the launch event, Baidu’s then Chief Operating Officer, Lu Qi, heralded the venture by saying, “AiBank is the future of intelligent finance… It is an institution that understands customers best and understands finance best.” This quote captures the aspiration that by fusing Baidu’s knowledge of users (from their online behavior) with Citic’s banking expertise, aiBank could outperform traditional banks in customer insight and service.

智能金融的未來…… 這是一間最了解客戶、同時最懂金融的機構。” 這句說話反映咗一個願景:通過結合百度對用戶(根據佢哋網上行為)嘅認知,配合中信嘅銀行專業,aiBank有機會喺客戶洞察同服務層面超越傳統銀行。

Being a direct bank (online-only) also meant aiBank could reach customers nationwide without physical presence, a significant advantage in China’s vast market.

作為一間直營(純網上)銀行,aiBank可以唔需要實體分行就做到全國覆蓋,呢個喺中國龐大市場入面係一個重要優勢。

In practice, over the next few years, aiBank rolled out digital lending products and AI-enhanced services. It offered personal loans via mobile apps, with quick approvals powered by machine learning credit models. For small businesses, it experimented with using AI to analyze e-commerce transactions and supply chain data to extend credit – much like Ant Group does.

實際上,喺隨後幾年,aiBank推出咗多種數碼貸款產品同AI加強服務。佢透過手機App提供個人貸款服務,用機器學習信貸模型快速批核。對於中小企,aiBank都試驗過用AI分析電商交易同供應鏈數據去擴展信貸——啱啱好就好似螞蟻集團嘅做法。

AiBank also explored AI in customer service, including intelligent chatbots for basic inquiries. Given Baidu’s strengths in natural language processing (Chinese-language NLP in particular), aiBank likely benefited from advanced AI in voice assistants and text-based customer interaction. While detailed performance data of aiBank is not widely public, its continued operation and capital increases (Citic and Baidu reportedly doubled its capital by 2018 to support growth ) suggest it gained traction.

aiBank亦都有探索AI喺客戶服務嘅應用,包括用智能聊天機械人解答基本查詢。因為百度喺自然語言處理(特別係中文NLP)具有領先優勢,aiBank好大機會受惠於先進AI語音助手同文字互動。雖然aiBank詳細業績數據唔係公開發佈,不過見到佢繼續營運同資本多次增持(有報道話中信同百度喺2018年已經為業務增資一倍支持增長),都顯示佢越做越有聲有色。

One unique angle for aiBank is the synergy with Baidu’s ecosystem. Baidu could integrate aiBank’s financial services into its popular apps. For instance, users of Baidu’s search or maps might be offered aiBank services contextually (imagine searching for “car loan” and seeing an aiBank offer). Moreover, Baidu’s AI research, such as in facial recognition and voice tech, found a real-world use in aiBank’s security and onboarding processes. As Yang Qiang from WeBank mentioned generally, technologies like facial recognition can allow seamless, remote account opening – aiBank likely employed similar methods given Baidu’s expertise. In a sense, aiBank served as a platform for Baidu to demonstrate the power of AI in a regulated industry, potentially strengthening Baidu’s position in the AI business market.

aiBank另一個獨特之處係同百度生態圈嘅協同。百度可以將aiBank金融服務整合入自己熱門App入面。例如,百度搜尋或者地圖用戶,喺相關情境下都可以即時見到aiBank服務(例如你搵“車貸”,就見到aiBank推介)。再加上百度AI科研(例如人臉識別同語音技術)都可以實際應用喺aiBank安全同開戶流程之中。正如微眾銀行楊強咁講,人臉識別等技術令開戶無縫遠程搞得掂——有百度專長,aiBank都應該用咗類似方案。某程度上,aiBank係百度展示AI喺受監管行業應用威力的平台,亦有助百度喺AI商業市場企穩陣腳。

However, running an AI-native bank within a larger traditional bank (Citic) structure also had challenges.

不過,要喺傳統銀行(中信)體系底下經營一間AI原生銀行,都有唔少挑戰。

Citic Bank’s involvement ensured regulatory compliance and provided banking infrastructure, but it may have also imposed a more cautious pace than a pure startup. Regulatory oversight by the China Banking and Insurance Regulatory Commission (CBIRC) meant aiBank’s AI innovations had to align with financial risk regulations. In 2021, an anecdote emerged that Chinese regulators fined Citic and Baidu for some formalities in the JV’s formation – a reminder that even tech-forward banks operate under strict rules. Nonetheless, China’s regulators have been generally supportive of AI and fintech in banking, as long as risks are controlled.

中信銀行嘅參與,確保咗aiBank有合規基礎同銀行基建,但係相比初創企業,可能都拖慢咗發展步伐。銀保監會(CBIRC)嘅監管意味住aiBank嘅AI創新要同金融風險條例接軌。2021年有消息指,內地監管機構曾經因合資公司成立過程中嘅某啲手續,向中信同百度罰款——都係提醒大家,即使係最前沿科技嘅銀行都要守規則。不過,只要風險受控,中國監管機構對銀行AI同金融科技一直都算支持。

As of 2025, CITIC aiBank stands as an example of a successful integration of AI in a new banking venture.

去到2025年,中信aiBank成為咗AI技術成功融入新型銀行業務嘅一個代表案例。

It may not have the global name recognition of WeBank, but it underscores a collaborative model: a legacy bank and a tech giant co-creating an AI-native banking platform.

雖然佢未必有微眾銀行咁國際知名,但都突出咗一種合作模式:即係傳統銀行同科技巨頭一齊打造AI原生銀行平台。

Closing Thoughts

The rise of native AI banks points to a future where finance is faster, more personalized, and even machine-driven.

AI原生銀行嘅崛起,預示住金融業未來會更快、更個性化,甚至由機器主導。

These pioneering projects demonstrate that banks can be radically rethought with modern technology – potentially offering customers ultra-convenient services and opening the financial system to new participants (like AI agents or underserved populations). Going forward, we can expect to see traditional banks respond by accelerating their own AI adoption or partnering with AI-native initiatives. In some cases, incumbents might acquire successful AI banking startups to bolt on their capabilities. Regulators, too, are paying close attention. If AI-native banks show strong performance in risk management and compliance, regulators may update frameworks to facilitate wider use of AI in banking, perhaps even creating new license categories for AI-driven financial institutions.

呢啲先鋒項目證明,利用現代科技銀行可以完全顛覆重塑——等客戶享受到極方便服務,亦令新參與者(如AI代理人或者長期被忽略群體)有機會加入金融系統。未來傳統銀行可能會加快採用AI,或者同AI原生機構合作。有啲情況下,老字號銀行可能會收購成功嘅AI銀行初創,增強自身實力。監管機構都密切留意住情況——如果AI原生銀行喺風險管理同合規做得好,監管層可能會更新規則,推動金融AI化,甚至設立AI金融機構專屬牌照。

However, the advent of AI-native banks also brings significant risks and challenges that need to be managed. One major concern is governance and oversight. When AI algorithms make credit decisions or detect fraud, ensuring they are unbiased and error-free is critical. Unchecked algorithms could inadvertently redline certain customer groups or approve risky loans – mistakes that could erode trust and invite regulatory penalties. Transparency is another challenge: these banks must make their AI’s actions explainable to regulators and customers.

當然,AI原生銀行出現同時都帶嚟唔少重大風險同挑戰。最大問題之一就係管治同監督。當用AI算法去批核貸款或者防止詐騙,點保證演算法無偏見、無重大錯誤係好關鍵。如果無好好監督,有可能錯誤排斥特定客戶群,或者批出高風險貸款——呢類失誤會損害信任,引嚟監管處罰。透明度都係大挑戰:AI銀行要令監管和客戶都明白演算法背後嘅決定點解咁做。

For traditional financial institutions, the emergence of AI-native banks is a double-edged sword. On one hand, it pushes the envelope of innovation, potentially yielding new methods and technologies that incumbents can adopt. Established banks can learn from the efficiency of Catena’s AI workflows or the customer engagement success of Bunq’s Finn, and integrate similar ideas. On the other hand, these new entrants could become formidable competitors in certain segments.

對於傳統金融機構嚟講,AI原生銀行冒起可謂一把雙刃劍。一方面,呢股新勢力推動創新,帶來新方法、新技術,畀老字號銀行可以借鏡——好似學Catena嘅AI流程效率,或者啟發自Bunq和Finn嘅客戶互動成效。另一方面,呢啲新玩家亦有機會喺某啲細分市場成為強勁對手。

免責聲明及風險提示: 本文資訊僅供教育與參考之用,並基於作者意見,並不構成金融、投資、法律或稅務建議。 加密貨幣資產具高度波動性並伴隨高風險,可能導致投資大幅虧損或全部損失,並非適合所有投資者。 文章內容僅代表作者觀點,不代表 Yellow、創辦人或管理層立場。 投資前請務必自行徹底研究(D.Y.O.R.),並諮詢持牌金融專業人士。
AI原生銀行:五大項目重塑金融行業 | Yellow.com