[Paper Review] Putting AI Ethics into Practice: The Hourglass Model of Organizational AI Governance

Matti Mäntymäki, Matti Minkkinen+2 more|arXiv (Cornell University)|Jun 1, 2022

Ethics and Social Impacts of AI31 citations

TL;DR

Introduces the hourglass model of organizational AI governance to translate AI ethics into practice, connecting environmental, organizational, and AI-system levels to align with the European AI Act throughout the system lifecycle.

ABSTRACT

The organizational use of artificial intelligence (AI) has rapidly spread across various sectors. Alongside the awareness of the benefits brought by AI, there is a growing consensus on the necessity of tackling the risks and potential harms, such as bias and discrimination, brought about by advanced AI technologies. A multitude of AI ethics principles have been proposed to tackle these risks, but the outlines of organizational processes and practices for ensuring socially responsible AI development are in a nascent state. To address the paucity of comprehensive governance models, we present an AI governance framework, the hourglass model of organizational AI governance, which targets organizations that develop and use AI systems. The framework is designed to help organizations deploying AI systems translate ethical AI principles into practice and align their AI systems and processes with the forthcoming European AI Act. The hourglass framework includes governance requirements at the environmental, organizational, and AI system levels. At the AI system level, we connect governance requirements to AI system life cycles to ensure governance throughout the system's life span. The governance model highlights the systemic nature of AI governance and opens new research avenues into its practical implementation, the mechanisms that connect different AI governance layers, and the dynamics between the AI governance actors. The model also offers a starting point for organizational decision-makers to consider the governance components needed to ensure social acceptability, mitigate risks, and realize the potential of AI.

Motivation & Objective

• Motivate addressing AI-related risks and harms (e.g., bias, discrimination) in organizational AI use.
• Provide a comprehensive governance framework to translate AI ethics principles into practice.
• Align organizational AI governance with forthcoming regulatory standards (EU AI Act).
• Offer a lifecycle-informed approach to governance that covers environmental, organizational, and AI-system levels.

Proposed method

• Proposes the hourglass governance framework as a structure for organizations that develop and use AI systems.
• Maps governance requirements to environmental, organizational, and AI-system levels.
• Links AI-system governance to the system life cycle to ensure governance across the life span of AI deployments.
• Highlights the systemic nature of AI governance and the need for mechanisms connecting governance layers.
• Suggests directions for practical implementation and future research on governance mechanisms and actor dynamics.

Experimental results

Research questions

• RQ1 How can ethical AI principles be translated into practical governance across an organization?
• RQ2 How do environmental, organizational, and AI-system levels interact within AI governance to ensure social acceptability and risk mitigation?
• RQ3 How can governance be aligned with existing or forthcoming regulations such as the European AI Act?
• RQ4 What mechanisms connect different AI governance layers and actors to support responsible AI throughout the system life cycle?

Key findings

• Introduces the hourglass model as a governance framework for organizational AI use.
• Emphasizes the systemic and multi-layered nature of AI governance across environmental, organizational, and AI-system levels.
• Connects governance requirements to AI-system life cycles to support governance throughout the system’s life span.
• Identifies new research avenues on practical implementation, inter-layer mechanisms, and actor dynamics in AI governance.
• Provides a starting point for decision-makers to consider governance components that realize social acceptability and risk mitigation.