Towards self-regulating AI: challenges and opportunities of AI model governance in financial services

  title={Towards self-regulating AI: challenges and opportunities of AI model governance in financial services},
  author={Eren Kurshan and Hongda Shen and Jiahao Chen},
  journal={Proceedings of the First ACM International Conference on AI in Finance},
  • E. Kurshan, H. Shen, Jiahao Chen
  • Published 2020
  • Computer Science, Economics, Mathematics
  • Proceedings of the First ACM International Conference on AI in Finance
AI systems have found a wide range of application areas in financial services. Their involvement in broader and increasingly critical decisions has escalated the need for compliance and effective model governance. Current governance practices have evolved from more traditional financial applications and modeling frameworks. They often struggle with the fundamental differences in AI characteristics such as uncertainty in the assumptions, and the lack of explicit programming. AI model governance… Expand

Figures from this paper

Mapping global AI governance: a nascent regime in a fragmented landscape
The rapid advances in the development and rollout of artificial intelligence (AI) technologies over the past years have triggered a frenzy of regulatory initiatives at various levels of governmentExpand
Seven challenges for harmonizing explainability requirements
It is argued that based on the current understanding of the field, the use of XAI techniques in practice necessitate a highly contextualized approach considering the specific needs of stakeholders for particular business applications. Expand
Beyond Fairness Metrics: Roadblocks and Challenges for Ethical AI in Practice
It is argued that a holistic consideration of ethics in the development and deployment of AI systems is necessary for building ethical AI in practice, and exhort researchers to consider the full operational contexts ofAI systems when assessing ethical risks. Expand
Ontology drift is a challenge for explainable data governance
We introduce the needs for explainable AI that arise from Standard No. 239 from the Basel Committee on Banking Standards (BCBS 239), which outlines 11 principles for effective risk data aggregationExpand
Graph Computing for Financial Crime and Fraud Detection: Trends, Challenges and Outlook
The threat landscape is analyzed and it is argued that it provides key insights for developing graph-based solutions and indicates emerging challenges in the continued effectiveness of the detection techniques. Expand


Transforming Paradigms: A Global AI in Financial Services Survey
This report presents the findings of a global survey on AI in Financial Services jointly conducted by the Cambridge Centre for Alternative Finance (CCAF) at the University of Cambridge Judge BusinessExpand
A Layered Model for AI Governance
This article proposes a conceptual framework for thinking about governance for AI, which aims to bridge the information gap between the developers of such systems and consumers and policymakers. Expand
Ensuring regulatory compliance in banking and finance through effective controls: The principle of duality in the segregation of duties
Today the segregation of duties is commonly used to ensure regulatory compliance in various industries. This article considers the organizational requirements for the effective implementation of thisExpand
Who Regulates Whom? An Overview of the U.S. Financial Regulatory Framework
This report attempts to set out the basic frameworks and principles underlying U.S. financial regulation and to give some historical context for the development of that system. The first sectionExpand
Algorithmic Decision Making and the Cost of Fairness
This work reformulate algorithmic fairness as constrained optimization: the objective is to maximize public safety while satisfying formal fairness constraints designed to reduce racial disparities, and also to human decision makers carrying out structured decision rules. Expand
Model-Agnostic Characterization of Fairness Trade-offs
This work hinges on the observation that many widely-used fairness definitions can be expressed via the fairness-confusion tensor, an object obtained by splitting the traditional confusion matrix according to protected data attributes, which yields a data-dependent yet model-agnostic way of understanding several types of trade-offs. Expand
Regulatory Sanctions and Reputational Damage in Financial Markets
We study the impact of the enforcement of financial regulation by the UK’s regulatory authorities on the market price of penalized firms. Existing studies rely on analyses of multiple events that mayExpand
Paradoxes in Fair Machine Learning
This work empirically evaluates the cost of allocation rules that satisfy both equalized odds and axioms of fair division on a dataset of FICO credit scores. Expand
Human-level control through deep reinforcement learning
This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks. Expand
Deep Learning in Finance
This work explores the use of deep learning hierarchical models for problems in nancial prediction and classication and finds that deep learning can detect and exploit interactions in the data that are, at least currently, invisible to any existing nancial economic theory. Expand