The Ethics of Artificial Intelligence for the Sustainable Development Goals

Contents
Contributors
Part I: AIxSDGs: Theory and Governance
	Introduction: Understanding the Ethics of Artificial Intelligence for the Sustainable Development Goals
		References
	AI in Support of the SDGs: Six Recurring Challenges and Related Opportunities Identified Through Use Cases
		1 Introduction
		2 Governance and Collaboration
			2.1 The Challenges
			2.2 The Opportunities
		3 Private Investments and the Role of Big Tech Companies
			3.1 The Challenges
			3.2 The Opportunities
		4 AI and Communities
			4.1 The Challenges
			4.2 The Opportunities
		5 AI and Individuals
			5.1 The Challenges
			5.2 The Opportunities
		6 Jobs and Skills
			6.1 The Challenges
			6.2 The Opportunities
		7 Impact Assessment
			7.1 The Challenges
			7.2 The Opportunities
		8 Evaluation and Limitations
		9 Concluding Remarks
		References
	Joined Up Thinking on How AI Can Contribute to the SDGs
		1 Background on AI for the SDGs and Changing Landscapes of R&D
		2 Misalignment of R&D/STI and the SDGs
		3 AI in the Context of Intelligence for the SDGs
		4 From Supply Push to Demand Pull
		5 Climate Change, Data and AI
		6 Conclusion
	Untitled
	A Realist’s Account of AI for SDGs: Power, Inequality and AI in Community
		1 Introduction
		2 AI for Sustainable Development: The Hope, the Hype and the Narrative
		3 Power, Inequality and Techno-solutionism
			3.1 Power
			3.2 Design and Deployment of AI and Social Inequality
		4 Tech Companies Heralding a New Digital Colonialism?
			4.1 Relationships of Dependency Affecting Vulnerable Communities
			4.2 Problem with AI Ethics
		5 Pathways to Sustainability
			5.1 Defining Innovation
			5.2 Theories and Systems Change
			5.3 The Role of AI in Solutions
		6 Digital Self-Determination
		7 AI in Community for Communal Empowerment
		8 Conclusion
		References
	The Potential of Artificial Intelligence for Achieving Healthy and Sustainable Societies
		1 Introduction
		2 Improved Health Through AI
			2.1 Shortage of Healthcare Workforce
			2.2 One Health and AI
			2.3 GeoAI for Precision Medicine
			2.4 Ethical and Societal Considerations
		3 Towards Sustainable Cities with the Help of AI
			3.1 AI for Extracting Climate-Related Indicators from Cities
				3.1.1 Mapping
				3.1.2 Predictive Models
				3.1.3 Generative Models
				3.1.4 Explainability
			3.2 Non-intrusive Evaluation of Air Quality in Urban Areas Through AI
			3.3 The Role of AI in Efficient and Sustainable Urban Mobility
		4 AI for Ambitious Climate-Action Targets
			4.1 The Potential Role of Artificial Intelligence to Combat Climate Change
			4.2 AI in Support of Understanding Climate Change
			4.3 AI in Support of Low-Carbon Energy Systems
			4.4 AI in Service of Energy Efficiency
			4.5 Engagement of AI on Climate Change
		5 Conclusions and Outlook
		References
	Artificial Intelligence: Poverty Alleviation, Healthcare, Education, and Reduced Inequalities in a Post-COVID World
		1 The UN SDGs
		2 A Brief Exploration of Various Theories of Poverty
		3 The Relevant Connection of AI to Poverty Alleviation
		4 Healthcare
		5 Education
		6 Conclusion
		References
	Missing Circles: A Dignitarian Approach to Doughnut Economics Through AI Applications
		1 Introduction
		2 The Doughnut Model
		3 Dignity, But Which One
			3.1 Traditional Conceptualisations of Human Dignity
			3.2 A Shift in Ethics: From Agents to Patients
			3.3 The Doughnut’s Allegiance
		4 The AI and the Doughnut
			4.1 Threatening the Boundaries
			4.2 The Missing Circles
		5 Conclusions
		References
	The Role of AI in SDG: An African Perspective
		1 Introduction
			1.1 AI and SDG
			1.2 AI and SDG 3: Current State in Africa Today
			1.3 AI and SDG 16: Current State in Africa Today
		2 Challenges of Integrating AI in the SDGs
		3 Discussions and Way Forward
		4 Conclusion
		References
	Artificial Intelligence for Advancing Sustainable Development Goals (SDGs): An Inclusive Democratized Low-Code Approach
		1 Introduction
		2 Research Problem and Research Questions
		3 Methods
			3.1 Rationale for Using the AI-Based Bayesian Network Approach
			3.2 The Bayesian Theorem
			3.3 The Research Model
		4 Discussion
			4.1 Example 1: An Inclusive and Democratized Low-Code Approach of Using AI for Global Sustainable Development
				4.1.1 Environmental Performance Index
				4.1.2 How Unified Analytics of Sustainability Indicators (Related to EPI and SDGI) Can Inform Education and Policymaking
			4.2 Example 2: An Inclusive and Democratized Low-Code Approach of Using AI for Ameliorating Malnutrition
			4.3 Example 3: An Inclusive and Democratized Low-Code Approach of Using AI for Financial Inclusion
			4.4 Example 4: An Inclusive and Democratized Low-Code Approach of Using AI for Improving Food Security
		5 Conclusion
		References
	Ethical AI: The European Approach to Achieving the SDGs Through AI
		1 AI Legal Framework in Europe: How Did Europe Get Here
		2 The Sustainable Development Goals and Their Connections with the Development of the AI Legal Framework
		3 Characteristics of the AI Regulation Proposal That May Foster the Achievement of SDGs
		4 Aspects of the AI Regulation Proposal That May Hinder the Completion of the SDGs
		5 Conclusions
		References
	AI as a SusTech Solution: Enabling AI and Other 4IR Technologies to Drive Sustainable Development Through Value Chains
		1 Challenge
		2 Solution
			2.1 Create a Sustainable Technology Board
				2.1.1 Provide a Platform for Cooperation
				2.1.2 Generate Analysis and Options
				2.1.3 Develop Standards and Guidelines
				2.1.4 Precedent and Practice
			2.2 Barriers and Solutions to SusTech Adoption
				2.2.1 Establish ‘Data Trusts’ to Share Data Safely and Securely
				2.2.2 Use Homomorphic Encryption to Share Data Safely and Securely
				2.2.3 Adopt a Typology for Data to Facilitate Management and Sharing
				2.2.4 Ensure Right-Skilling Programmes Match Skills Supply to Skills Demand
				2.2.5 Develop Innovative Technology Finance (TechFin) Instruments
				2.2.6 Orient Investment Incentives to Encourage the Uptake of SusTech Solutions
				2.2.7 Incorporate Non-equity Modes or Strategic Partnerships in Domestic and International Policy Frameworks
				2.2.8 Use Performance-Based Regulation to Balance Flexibility with Oversight
				2.2.9 Use Sustainability Impact Assessments
				2.2.10 Ensure Equivalency Agreements on Standards and Certifications
				2.2.11 Build Living Labs and (International) Regulatory Sandboxes
			2.3 Examples of SusTech in Action
				2.3.1 Artificial Intelligence
				2.3.2 Blockchain
				2.3.3 Internet of Things
				2.3.4 Automation and Drones
				2.3.5 Augmented and Virtual Reality
		3 Conclusion
		References
	Untitled
	Untitled
	AI for Sustainable Finance: Governance Mechanisms for Institutional and Societal Approaches
		1 Introduction
		2 AI and the SDGs
		3 The Promise of AI for Sustainable Finance
			3.1 A Brief History
			3.2 Recent Governance Responses
		4 Institutional and Societal Approaches
			4.1 ESG Investing
				4.1.1 Related Governance Challenges
			4.2 Financial Inclusion
				4.2.1 Related Governance Challenges
		5 Navigating Through a Pandemic
			5.1 COVID-19: A “Natural Experiment”
				5.1.1 Related Governance Challenges
		6 Key Policy Considerations
			6.1 Mitigate Unintended Social and Environmental Consequences
			6.2 Promote ESG Disclosure
			6.3 Strengthen Cross-Sector Partnerships
		7 Conclusion
		Works Cited
	Big Tech Corporations and AI: A Social License to Operate and Multi-Stakeholder Partnerships in the Digital Age
		1 Introduction
		2 UNs SDGs Framework and Its Link with AI Challenges and Impacts
			2.1 The When and Why of the UN SDGs
			2.2 AI for Social Good
		3 Towards Sustainable Digital Business Models: Some Reflections on the Co-presence of Different Spheres and Values
		4 The Need for a ‘Social License to Operate’ in the Digital Age
		5 Conclusion
		References
Part II: AIxSDGs: Existing and Potential Use Cases
	A Legal Identity for All Through Artificial Intelligence: Benefits and Drawbacks in Using AI Algorithms to Accomplish SDG 16.9
		1 Introduction
		2 The Concept of Identity: External Variables and Internal Dimension
			2.1 Different Societies, Different Identities
		3 Artificial Intelligence and Identity
			3.1 AI-Based Identification Procedures in the Migration Context
			3.2 Artificial Intelligence for Identification Purposes: The EU Experience
		4 AI-Based Devices for Identification Purposes: Ethical Concerns and Legal Issues
			4.1 Artificial Intelligence and the Principle of Non-discrimination in the Context of Identification Operations
			4.2 Artificial Intelligence, Privacy Rights and Identity Issues
		5 Concluding Remarks and Recommendations
		References
	Socially Good AI Contributions for the Implementation of Sustainable Development in Mountain Communities Through an Inclusive Student-Engaged Learning Model
		1 Introduction
		2 History and Background
		3 Why Focus on Sustainable Development in Mountains?
			3.1 Mountain-Focused IT and IS Specialization Initiatives
			3.2 Contributions of Education to the Implementation of the Agenda Based on Socially Good Principles
		4 Goals and Targets Related to SMD
			4.1 Inclusive Student-Engaged Learning as a Foundation for a Socially Good Implementation of SMD
			4.2 Examples of Socially Good SMD Advocacy and IoT Use Within the Adapted SEL Model
		5 Recommendations and Conclusion
		Bibliography
	Gender, Health, and AI: How Using AI to Empower Women Could Positively Impact the Sustainable Development Goals
		1 Introduction
		2 AI, Medicine, and Gender
			2.1 State of the Art: AI in Medicine
			2.2 AI in Practice: How Do AIs Work?
			2.3 Use Cases: Where Are All the Women Gone?
		3 Gender-Balanced AI for SDG
			3.1 Gender-Balanced AI and Gender Equality as a Part of Sustainable Development: Focus on SDG 5
			3.2 Impact on Other SDGs: Health, Economics, Innovation, and Inequalities—SDGs 3, 8, 9, and 10
		4 Conclusions and Future Research
		Bibliography
	Smart Control of Drinking Water Grids Using IoT
		1 Introduction
		2 Related Works
			2.1 Water Quality Monitoring Systems
			2.2 Leak Detection Solutions in the Water Distribution Network
		3 The Proposed System Architecture
		4 A New Model for Water Quality Analysis Based on Machine Learning
			4.1 Data Gathering
			4.2 Data Aggregation
			4.3 Classification with Machine Learning
				4.3.1 Decision Tree Algorithms
				4.3.2 Support Vector Machine
				4.3.3 KNN
		5 The Proposed Leak Detection Algorithms in the Water Distribution Network
			5.1 The Small Leaks Control Algorithm
			5.2 The Large Leaks Control Algorithm
		6 Experimentation
			6.1 Water Quality Evaluation
				6.1.1 Accuracy Evaluation
				6.1.2 Precision Evaluation
				6.1.3 Recall Evaluation
			6.2 Leak Detection Evaluation
				6.2.1 Evaluation of the Water Pressure Evolution in Adjacent Pipes
				6.2.2 Leak Detection Evaluation
		7 Conclusion and Perspectives
		References
	Algorithmic Art and Cultural Sustainability in the Museum Sector
		1 Introduction: Cultural Datasets, Cultural Algorithms
		2 Sustainable Development and AI-Based Technologies
		3 Cultural Sustainability and the Museum Space
		4 An Algorithmic Art Framework for Techno-cultural Sustainability
		5 Conclusion
		References
	The Impact of Artificial Intelligence on Circular Value Creation for Sustainable Development Goals
		1 Introduction
		2 AI in CE
		3 Circular Value Creation
		4 AI and Circular Value Creation
		5 Discussions
		6 Conclusion
		References
	Computer-Aided Corporate Sense-Making and Prioritization for SDGs
		1 Introduction
		2 Motivation
		3 Proposed Methodology
		4 Lessons Learned
			4.1 Minimizing Respondent Burden
			4.2 Assessments About the Present or the Future?
			4.3 Why Are the Company Strategy and Initiative Impact Assessments (So) Different?
			4.4 The Kind and Direction of the SDG Impact
			4.5 The Ethics of Computer-Aided Minimization of Miscommunications
			4.6 Role Conflicts of Respondents
		5 Conclusion
		References
	Role of Artificial Intelligence in Advancing Sustainable Development Goals in the Agriculture Sector
		1 The Ever-Growing Hunger
		2 Looking Towards AI to Solve Agricultural Problems
		3 Overview of Data Sources That Are Being Collected Across the Agricultural Chain
		4 Leveraging AI for Advanced Agricultural Outputs
			4.1 Selection of High Resistance Variety of Crops
			4.2 Working and Learning with Advanced Data
			4.3 Precision Agriculture
			4.4 Augmenting Labour Force and Skills
			4.5 Maximising Returns
			4.6 Chatbots for Farmers
			4.7 Intelligent Crop Planning
			4.8 Postharvest Value Chain Operations
		5 Can AI Impede Achievement of SDGs in Agriculture?
			5.1 Need for Massive Computational Resources
			5.2 Potential to Accelerate Inequalities
			5.3 Uneven Distribution of AI Systems
		6 Challenges with AI Adoption in Agriculture
			6.1 Structured and Coherent Data
			6.2 Lack of Knowledge
			6.3 Limited Scope of Scalability
			6.4 Poor Awareness of the Farm Production Functions
			6.5 Technological Infrastructure and Investment
		7 Conclusion
			7.1 Strengthening of Skills and Capacities
			7.2 Cooperation and Readiness Amongst Key Stakeholders
			7.3 Mitigation of Data and Infrastructure-Related Risks
			7.4 Readiness of the Key Stakeholders
		References
	Untitled
	Untitled
	Untitled
	AI for Sustainable Agriculture and Rangeland Monitoring
		1 Introduction
			1.1 Problem Overview
			1.2 AI-EO SDG Model: Zero Hunger
			1.3 AI-EO SDG Model: Climate Change
		2 Background
			2.1 Rangelands
			2.2 AI for Climate Change and Agriculture: Overview
		3 Methods
			3.1 Overview of the Proposed Approach
			3.2 Data
			3.3 Model Architecture
				3.3.1 Habitat Classification
				3.3.2 Biomonitoring Parameter Estimation
				3.3.3 AI Model Results
				3.3.4 Decision Support Model
		4 Discussion
			4.1 Challenges and Economic Implications
			4.2 Economic Outcomes
		References
	Artificial Neural Networks Predict Sustainable Development Goals Index
		1 Introduction
		2 Sustainable Development Goals Index (SDGI)
		3 Artificial Neural Networks (ANNs)
		4 Genetic Algorithm (GA)
		5 Method
		6 Results
		7 Conclusions
		References
	Sailing the Data Sea to Advance Research on the Sustainable Development Goals
		1 Introduction
		2 Information Extraction
		3 Data
		4 Methodology
			4.1 Parsing Papers
			4.2 Binary Text Classification
			4.3 Named Entity Recognition (NER)
				4.3.1 Bidirectional LSTM with CRF
				4.3.2 BERT
			4.4 Classification by SDGs
		5 Results and Discussion
			5.1 Parsing Papers
			5.2 Binary Text Classification
			5.3 Named Entity Recognition
			5.4 Classification by SDGs
			5.5 Discussion
		6 Conclusion and Future Work
		References
	An Empirical Analysis of AI Contributions to Sustainable Cities (SDG 11)
		1 Introduction
		2 Related Work
			2.1 AI for SDG 11
			2.2 Citizen-Centric Approach for SDG 11
			2.3 Exiting Gaps
		3 Method
		4 Results
			4.1 Geographic Distribution of AI Projects
			4.2 Key Beneficiaries of AI Projects
			4.3 Types of Systems
			4.4 Targets Served by AI Projects
			4.5 Indicators Supported by AI Projects
			4.6 Contribution to Citizen-Centric Challenges
		5 Discussion
			5.1 Key Takeaways
			5.2 Limitations
			5.3 Future Work
		6 Conclusion
		Appendices
			Appendix A: Definition of the SDG 11 Targets Found in the Dataset
			Appendix B: Definition of the SDG 11 Indicators Found in the Dataset
		References
Index