On 25 July 2019, the Dutch Central Bank (De Nederlandsche Bank, DNB) published a guidance document containing general principles for the use of artificial intelligence (AI) in the financial sector (the Guidance Document). The Guidance Document serves as a discussion paper containing DNB’s preliminary views on the responsible use of AI in the financial sector. DNB believe that the issues and ideas outlined in this paper will benefit from broader debate, and therefore DNB welcomes comments on the discussion paper.
According to DNB, financial undertakings increasingly make use of AI to enhance their business processes, and improve their product and service offerings. Examples of current AI applications are identity verification in customer on-boarding, transaction data analysis, fraud detection in claims management, pricing in bond trading, automated analysis of legal documents, customer relation management, and risk management. Although AI enables these undertakings to enhance their business processes and provide new added value, at the same time incidents with AI could harm a financial undertaking or its customers and can have serious reputation effects for the financial system as a whole. For this reason, financial undertakings using AI should adhere to principles of sound and controlled business operations. DNB believes that a responsible use of AI in financial services entails that financial undertakings should pay due attention to the soundness, accountability, fairness, ethics, skills and transparency aspects of the AI applications that they develop.
In the Guidance Document, DNB has formulated a number of general principles regarding the use of AI in the financial sector divided over six key aspects of responsible use of AI: soundness, accountability, fairness, ethics, skills and transparency.
Important points noted by the DNB in the Guidance Document:
- AI applications in the financial sector need to be reliable and accurate, behave predictably, and operate within the boundaries of applicable rules and regulations;
- financial undertakings need to be accountable for the use of AI applications (AI applications that do not function right could result in damages for the financial undertaking, its customers and other stakeholders);
- AI applications should not inadvertently disadvantage certain groups of customers;
- all people in a financial undertaking (work floor to the board room) should have a sufficient understanding of the strengths and limitations of the AI-enabled systems that they work with; and
- financial undertakings need to be able to explain how and why they use AI in their business processes and (where reasonably appropriate) how these applications function.
DNB emphasises that as AI applications increasingly inform a financial undertaking’s decisions and as their potential consequences for the financial undertaking and its customers grow, responsibility and accountability standards governing their deployment will become stricter.
View the Guidance Document, 25 July 2019.