Overseas Leading Financial Firms Address Internal Control Issues with AI and Big Data

Woori Financial Research Institute's 'Digital Internal Control System Case'
Advanced Adoption of Regtech
Applied to Anti-Money Laundering, Credit, and Marketing

Wall Street Sign (Asia Economy = Yonhap News)

DBS Bank Singapore's anti-money laundering (AML) platform, 'Cruise,' detects suspicious transactions from countless financial transaction data and assesses the risk level of alerts as 'High, Medium, Low, Very Low.' This prioritization helps prevent unnecessary investigations. It also provides visualized analytical data linking related suspicious transactions and customers' connections, enabling staff to grasp money laundering-related transactions at a glance.

Not Just Simple Risk Detection but Also Risk Level Classification

Overseas financial institutions are increasingly adopting artificial intelligence (AI) and big data as internal control measures to prevent financial incidents, like DBS Bank. The report titled 'Digital Internal Control System Cases,' released on the 27th by Woori Financial Research Institute, introduced advanced Regtech adoption cases. Regtech, a combination of Regulation and Technology, is used in anti-money laundering, credit, and marketing operations.

Before implementing 'Cruise' two years ago, DBS Bank experienced trial and error. Although AI and big data were used to select suspicious transactions, normal transactions meeting specific conditions such as large cash amounts, foreign exchange transactions, or repetitive transactions were also considered suspicious. These cases accounted for 98% of all alerts, resulting in time and manpower spent on unnecessary investigations.

The report stated, "With Cruise, risk levels are determined to enable early identification of money laundering financial incidents and rapid detection of illegal activities," adding, "The volume of suspicious transactions analyzed within the same time increased by 33%."

Applied to Marketing to Prevent Mis-selling

There is also a case where AI automatic inspection functions were introduced in marketing to prevent incidents like financial product mis-selling. Fidelity Asset Management in the United States launched the marketing compliance platform 'Saifr' to strengthen advertising review procedures. It reviews content by judging whether there are inappropriate expressions in advertisements and suggests alternative expressions along with reasons. Then, real-time exchanges of opinions occur among marketing, compliance, legal, and product development departments before modifications and approvals are finalized.

The report said, "Saifr interprets advertisements in text, image, and video formats and evaluates misleading, exaggerated, unfair, or unjust expressions as risks," adding, "The AI was trained on millions of documents worked on by in-house marketing and compliance departments over 15 years."

JP Morgan introduced the AI-powered contract review software 'COiN' to reduce errors in loan contract reviews. Since about 80% of loan incidents occur due to contract interpretation errors, improving document review accuracy is crucial.

The report explained, "It can detect special loan conditions that humans might easily overlook, such as collateral setting conditions or interest rate changes, classifies clauses into 150 attributes based on wording or sentence position in the contract, and highlights items requiring attention," adding, "The processing speed has also been reduced from 360,000 hours annually to just a few seconds."

Greater Responsibility for Internal Control, Focus on Prevention

Korean commercial banks are also applying AI and big data technologies to detect abnormal transactions, protect consumers, and prevent money laundering. The report emphasized, "It is necessary to advance systems to avoid inefficiencies caused by outdated technology and fragmented data, as seen in DBS Bank," and "Especially for customer-facing operations with significant potential harm, such as credit and product sales, adding AI specialized in contract review for these tasks is essential to prevent financial incidents."

Korean financial institutions now bear heavier responsibilities for internal control. On the 22nd, the Financial Services Commission required financial groups and commercial banks to create a 'responsibility structure chart' to prevent internal incidents and accidents. The core is to predefine internal control responsibility areas for 20 to 30 key executives, including the CEO. The financial authorities can sanction these executives accordingly. Consequently, a trend toward focusing on preventive internal control is expected to form among Korean banks.