近年各家銀行紛紛進行數位轉型，因應金融科技(Financial technology，簡稱 FinTech)及 COVID-19 疫情的擾動，催化金融機構除了在現行網路銀行增加更多便捷性功能，另外在傳統放款方面，也衍生出新金融模式—智能放款。它結合了人工智能、大數據及機器學習等技術，透過應用 AI 人工智能技術和大數據
分析判斷，實現了自動化 SOP、標準化 SOP、效率化及低成本的貸款流程。本研究旨在探討智能放款對金融機構授信風險的變動影響及後續發展，採用金融同業訪談方式蒐集資料，受訪之三家銀行佔台灣金控銀行及非金控銀行，放款餘額佔比約 8.44%，具有一定影響性。本研究透過訪談，了解智能放款對金融機構授信風險的影響、智能放款對金融機構風險管理策略的影響及如何應對智能放款對金融機構授信風險影響，提供現行金融機構做法進行分析。本研究之結論得出智能放款對金融機構授信風險的影響，並提出相應的建議。智能放款的風險控管能力較人工授信更佳，因為智能放款能夠快速分析大量資料，提高評估的準確性，同時避免由於人為因素而造成的偏見和錯誤判斷僅較仰賴資料準確性和模型穩定性的基礎上。研究過程中，涉及訪談者 該機構內部機密資訊，受訪者回答的內容則較為隱晦，雖得知數位轉型仍是各家銀行的當務之及，但面對風險的考量，包括資訊不對稱、冒貸風險、市場波動風險及個資保密風險等問題，尚需結合更多金融數據的串聯，成立專責部門及相關因應單位，以及主管機關對於金融法規的推動，使智能放款的數位轉型逐漸完善。

In recent years, various banks have carried out digital transformation one after
another. In response to financial technology (Financial technology, referred to as
FinTech) and the disturbance of the COVID-19 epidemic, financial institutions have
been catalyzed to add more convenient functions to the current online banking, and in
addition to traditional lending. Derived a new financial model - smart lending. It
combines technologies such as artificial intelligence, big data, and machine learning. Through the application of AI artificial intelligence technology and big data analysis and judgment, it realizes automated SOP, standardized SOP, efficient and low-cost loan process. The purpose of this study is to explore the impact of intelligent lending on the credit risk of financial institutions and its subsequent development. The data is collected by interviews with financial peers. The three banks account for Taiwan's financial holding banks and non-financial holding banks, and their loan balance accounts for about 8.44%. Influence, through interviews to understand the impact of smart lending on the credit risk of financial institutions, the impact of smart lending on the risk management strategies of financial institutions and how to deal with the impact of smart lending on the credit risk of financial institutions, provide the current financial institutions for analysis, and conclude that smart lending Make corresponding suggestions on the conclusion of the impact on the credit risk of financial institutions. The risk control ability of smart lending is better than that of manual credit extension, because smart lending can quickly analyze a large amount of data, improve the accuracy of evaluation, and avoid bias and misjudgment caused by human factors. Only rely on data accuracy and model stability basis. During the research process, interviewees were involved in the organization’s internal confidential information, and the content of the interviewees’ answers was relatively cryptic. Although they learned that digital
transformation is still a top priority for all banks, the risk considerations, including information asymmetry , Risk of risky loans, risk of market fluctuations, and risk of confidentiality of personal information. It is still necessary to combine more financial data in series, establish a dedicated department and related response units, and the promotion of financial regulations by the competent authority to make the digital transformation of smart loans gradually. Complete.

壹、 中文部份
張鼎煥、陳璦玲(2017)，「銀行住宅貸款授信品質研究-預期違約機率之應用」，明
道學術論壇 10(3), 01-18
楊茜文(2022)，「銀行不動產抵押貸款與貨幣傳遞機制:流動性與價格風險的實證
研究」，經濟論文叢刊, 50(4), 423-459
胡韶雯(2022)，「永續金融-後疫情時代責任授信法制研究」，國科會專題研究計畫
(MOST 110-2410-H-031-013-)
蔡翼擎、吳子謙、吳奕榮、陳浩、許瑋澄、李穎奕(2019)，「台灣金融服務業如何
運用 FinTech 進行轉型革命」，東亞論壇季刊, 503, 1-16
王仁聖、林冠仲(2019)，「金融科技(FinTech)商業模式策略優化研究:以法遵科技
(RegTech)為例」，科技管理學刊, 24(2), 1-30
陸璐(2020)，「”FinTech”賦能:科技金融法律規制的范式轉移」，政法論壇, 2020(1),
137-148
貳、 英文部份
Calhoun, C. A., & Deng, Y. (2002). A dynamic analysis of fixed- and adjustable-rate
mortgage terminations. Journal of Real Estate Finance and Economics, 24(1/2),
09-33.
Von Furstenberg, G. M. (1969). Default risk on FHA-insured home mortgages as a
function of the terms of financing: a quantitative analysis. Journal of Finance,
24(3), 459-477
Xu, N., & Wang, K.J.(2019). Adopting robot lawyer? The extending artificial
intelligence robot lawyer technology acceptance model for legal industry by an
exploratory study. Journal of Management & Organization 27(5), 867-885