The Financial Supervisory Commission (hereinafter referred to as the FSC) today released the “Guidelines for the Use of Artificial Intelligence (AI) in the Financial Industry,” aiming to encourage financial institutions to make good use of technology and apply trusted AI to develop financial services that are more responsive to the needs of the public.
Table of Contents:
Toggle
Reasons for FSC’s promotion of AI in the financial industry
Guidelines for the use of AI in the financial industry
General provisions
Six chapters
Key points of the guidelines
Chapter 1: Risk management
Chapter 2: Use of generative AI
Chapter 3: Customer privacy protection
Chapter 4: System interpretability and transparency
Administrative guidance nature
On October 17, 2023, the FSC released the “Core Principles and Related Promotion Policies for the Use of Artificial Intelligence (AI) in the Financial Industry,” and subsequently drafted the guidelines based on these core principles. During this period, the FSC referred to relevant guidelines issued by other national governments and international organizations, and solicited opinions from various sectors and experts through public policy participation platforms, fintech-related conferences, and secondary AI guideline consultation meetings. After multiple adjustments, the “Guidelines for the Use of AI in the Financial Industry” was finally released today as a reference for financial institutions to introduce, use, and manage AI.
The “Guidelines for the Use of AI in the Financial Industry” released this time mainly consists of general provisions and six chapters:
Advertisements – Please scroll down for the rest of the article
General provisions mainly explain common matters such as AI-related definitions, AI system life cycle, factors to consider in risk assessment, ways to implement core principles based on risk, and supervision and management of third-party providers.
The six chapters respectively explain the key points that financial institutions should focus on and the measures they can take in implementing the six core principles based on the AI life cycle and assessed risks, including purposes, main concepts, corresponding precautions for each principle, and implementation methods or measures.
Risk management of financial institutions may involve external reviews. The guidelines specify that financial institutions may establish mechanisms for review and evaluation conducted by independent third parties with AI expertise based on their own expertise and resource levels, as needed.
Considering that financial institutions may not necessarily have control over the results generated by generative AI, but in order to ensure fairness and human control of AI systems, the guidelines require financial institutions to implement risk controls on generative AI developed by third-party providers.
The guidelines reference the General Data Protection Regulation (GDPR) of the European Union and require financial institutions to adopt the “data minimization principle” to avoid collecting excessive or unnecessary information.
Financial institutions should ensure the interpretability of AI systems. However, considering that if they outsource the development or purchase of AI systems to other providers, they may not have complete knowledge of the operational details of the systems due to business confidentiality. Therefore, “interpretability” is limited to financial institutions clearly explaining how “AI systems developed and used by themselves or commissioned” operate and the logic behind their predictions or decision-making processes. In addition, to enhance market trust in AI systems, financial institutions can proactively disclose relevant information through reports, technical documents, or websites to make stakeholders aware of their use of AI systems.
The FSC emphasizes that these guidelines mainly provide matters to be noted by financial industry players when introducing and using AI and are of an administrative guidance nature. Financial industry-related associations can refer to these guidelines and incorporate relevant key points and measures into their self-disciplinary regulations if they have established guidelines for the use of AI. If no relevant self-disciplinary regulations have been formulated, financial institutions are recommended to refer to these guidelines for the introduction, use, and management of AI systems.
AI
FSC
Further reading:
AI tokens rise with Nvidia, AGIX, RNDR outperforming Bitcoin
Cryptocurrency + AI, how to leverage blockchain technology in the AI field?