What does Korea FSS AI Risk Management Framework say about AI in banking?

Published Jan 15, 2026 · Last reviewed Nov 20, 2026

South Korea's Financial Supervisory Service published its AI Risk Management Framework in the Financial Sector (AI RMF) in draft on 15 January 2026, setting out three core processes for financial companies: AI governance, AI risk assessment and AI risk control. It puts into practice the governance principle of the FSC's revised AI guidelines: a decision-making body such as an AI ethics or risk management committee, a risk unit independent of the teams that build AI, a risk grade for each AI service, and automatic high-risk treatment for high-impact AI under the AI Basic Act. The framework is voluntary; the FSC said it would be distributed with the revised guidelines when they took effect on 22 June 2026.

OFFICIAL TEXT: fss.or.kr ↗ · IN FORCE · SOUTH KOREA (FSC / FSS)

Checked 2026-11-20 · South Korea (FSC / FSS), fss.or.kr
DocumentKorea FSS AI Risk Management Framework — AI Risk Management Framework in the Financial Sector (금융분야 AI 위험관리 프레임워크, AI RMF)
Issued byFinancial Services Commission (FSC) and Financial Supervisory Service (FSS) of the Republic of Korea, with the Ministry of Science and ICT under the Framework Act on the Development of Artificial Intelligence and the Creation of a Foundation for Trust (AI Basic Act)
TypeFramework
StatusIn force
PublishedJan 15, 2026
EffectiveJun 22, 2026
Applies toFinancial companies supervised by the FSS that develop or use AI. Voluntary: the FSS says the framework has no legal force and can be strengthened, relaxed or applied selectively according to a company's size, resources, business, scope of AI use and technical maturity. Published in Korean; the English title is the one used in the FSC's 2026 guidelines. The linked FSS release of 15 January 2026 carries the draft, which the FSS then planned to finalise in the first quarter of 2026; the final text was not found on fss.or.kr, and its June 2026 effect rests on the FSC's 18 June 2026 statement that the FSS would provide the framework with the guidelines taking effect on 22 June
Official sourcefss.or.kr ↗
Use casesAI governance (general) · Model risk management · Credit scoring & underwriting · Third-party & vendor AI · Data & privacy

What are the key points of Korea FSS AI Risk Management Framework?

  • Starting point: an FSS survey of 118 companies (20 banks, 37 securities firms, 53 insurers and 8 card companies) in April 2025 counted 653 AI services, 299 of them at banks, and found that only 5 banks (25%), 4 insurers (7.5%) and 1 securities firm (2.7%) had set up an AI decision-making body, that about 85% had no AI ethics principles or risk management standards, and that about 60% applied the supervisors' AI guidelines.
  • Governance: a decision-making body (for example an AI ethics committee or risk management committee) deliberates on AI internal rules, risk management and consumer protection policies and approval of high-risk AI services, and its chair reports regularly to the CEO; companies that have adopted a responsibilities map are asked to consider writing AI duties into it.
  • Independent risk unit: a dedicated unit separate from the teams that introduce or develop AI oversees AI risk identification, measurement and evaluation and compliance with the AI Basic Act and other laws.
  • Internal rules: AI ethics standards, AI risk management rules and manuals covering high-risk AI controls, the AI life cycle, security, consumer protection and breaches of law, with final decisions and responsibility kept with staff.
  • Risk assessment: a risk-based approach (a rule-based approach is also allowed) that runs from risk identification and measurement through mitigation and residual risk to a risk grade for each AI service; the FSS example weights legality 20%, credibility 30%, good faith 20% and security 30%, scoring items such as bias, fairness, explainability, outsourcing and privacy.
  • High-impact AI: to stay consistent with the AI Basic Act, AI that is high-impact under that Act, whose examples include loan screening, is classed as high-risk whatever its score.
  • Risk control: lighter approvals and documentation for low-risk AI; prior approval by the AI ethics committee, post-launch verification, third-party evaluation and closer monitoring for high-risk AI; and re-examination of whether to launch very high-risk AI where the decision-making body sees a threat to financial stability or consumer rights. Monitoring, documentation, training and information sharing with supervisors complete the process.
  • Follow-up: the FSS said it would spread best practices and check implementation at adopting companies so that the processes take root in their internal controls.

What did Korea FSS AI Risk Management Framework change for banks?

Korea's earlier financial AI guidance set principles but no common governance and risk process, and the FSS survey found few companies had one. The framework gives firms a template that links each step to the FSC's seven principles and to the AI Basic Act's high-impact category. It was announced in draft for industry comment, and the FSC said it would be distributed with the revised AI guidelines when they took effect on 22 June 2026.

Is the FSS AI risk management framework mandatory?

No. The FSS describes it as a voluntary guideline without legal force that each company can strengthen, relax or apply selectively according to its size, resources, business, scope of AI use and technical maturity. Duties under the AI Basic Act, which the framework reflects, are binding separately.

How does the FSS AI RMF classify AI risk?

Each AI service is scored on factors drawn from the FSC's principles of legality, credibility, good faith and security; mitigation and residual risk are then assessed to set a risk grade that decides how strict the controls are. AI that is high-impact under the AI Basic Act, whose examples include loan screening, is always treated as high-risk.

DateDocumentStatus
Jul 2, 2026Korea FSC Frontier AI Security Response Guide — Financial Sector Response Guide to Frontier AI Security Threats (프런티어 AI 보안위협 금융분야 대응요령)Final
Jun 18, 2026Korea FSC AI Guidelines for the Financial Sector — AI Guidelines in the Financial Sector (금융분야 인공지능 가이드라인)In force
May 22, 2026Korea FSC Frontier AI Network Separation Easing — Network Separation Rules to be Eased in Financial Sector to Boost Innovation and Cybersecurity in AI TransformationFinal
Jan 21, 2025Korea AI Basic Act — Framework Act on the Development of Artificial Intelligence and the Creation of a Foundation for Trust (인공지능 발전과 신뢰 기반 조성 등에 관한 기본법)In force
Aug 13, 2024Korea FSC Network Separation Roadmap — Roadmap to Improve Network Separation in the Financial Industry (금융분야 망분리 개선 로드맵)Final
Apr 17, 2023Korea FSC AI Credit Scoring and AI Security Guideline — Verification System for AI-driven Credit Scoring Models and Security Guideline for the Use of AI in the Financial SectorFinal

Follow every document these regulators publish

when one of these regulators moves, the next morning's brief says so · six sourced stories · 7 am PT · free

plus every regulation, bank and agent page update, the morning after · leave any morning