Report on the Intersection of Artificial Intelligence and Sustainable Development Goal Research

Executive Summary

This report analyzes the application of Artificial Intelligence (AI) in research related to the United Nations Sustainable Development Goals (SDGs). Based on a review of 792 peer-reviewed articles, this analysis explores the current landscape, identifying trends, applications, and critical gaps. The findings indicate that while AI, particularly deep learning and supervised machine learning, is increasingly used for forecasting and optimization, its application is heavily concentrated in specific domains. A significant disconnect exists between advanced AI methodologies and deep sustainability expertise, hindering the full realization of AI’s potential to drive transformative change for the 2030 Agenda. The report highlights an urgent need to bridge this gap, especially in underrepresented social sustainability areas, to ensure AI contributes effectively and equitably to achieving all 17 SDGs.

1. Introduction: AI in the Context of the 2030 Agenda for Sustainable Development

The 2030 Agenda for Sustainable Development, with its 17 interlinked Sustainable Development Goals (SDGs), provides a global framework for a sustainable future. Achieving these goals requires integrated, systems-level solutions. Artificial Intelligence (AI) is positioned as a powerful tool to accelerate progress by enabling data-driven insights and optimizing complex systems. However, the rapid, uneven development of AI has outpaced the establishment of ethical and regulatory frameworks, raising questions about its true transformative potential for the SDGs. This report assesses the extent to which AI tools are being integrated with deep knowledge on sustainable development by reviewing highly cited, peer-reviewed literature at the intersection of AI and the SDGs.

2. Methodology

A systematic content analysis was conducted on research concerning AI and the SDGs. The methodology involved the following steps:

1. Data Collection: Metadata was extracted from the Scopus database in January 2024 for each of the 17 SDGs. Search criteria included terms for “Artificial Intelligence,” “sustainability,” and SDG-specific vocabulary.
2. Screening and Selection: From an initial pool of 14,423 articles, the most cited articles per SDG were screened. A final corpus of 792 articles that substantively engaged with both AI and one or more SDGs was selected for full-text analysis. SDG 17 (Partnerships for the Goals) was excluded due to a lack of relevant articles found.
3. Data Coding and Analysis: Articles were classified as either empirical (n=393) or conceptual/review (n=399). A multivariate statistical full-text analysis was performed on the empirical articles to identify thematic patterns and research clusters.

3. Results: Current State of AI Application in SDG Research

3.1. Overview of Publication Trends and Geographic Focus

The volume of research applying AI to SDGs has increased substantially, surpassing 200 publications annually in 2022 and 2023. The primary research output originates from institutions in Europe and Asia, with China, India, the United States, and Spain collectively authoring 38% of the analyzed articles.

A clear geographic variability exists in the research focus on specific SDGs:

• SDG 6 (Clean Water and Sanitation): Predominantly addressed by studies from Iran, India, and Spain.
• SDG 3 (Good Health and Well-being): A primary focus for researchers in Italy and the United Kingdom.
• SDG 4 (Quality Education): Most commonly researched by authors in the United States, Spain, and China.

3.2. Thematic Patterns in Empirical Studies

Hierarchical cluster analysis of the 393 empirical articles revealed eight distinct research groups organized along two primary axes:

1. Disciplinary Axis: Ranging from natural sciences (e.g., hydrological systems, vegetative assessments) to the humanities (e.g., health, education).
2. Focus Axis: Distinguishing between studies with an economic focus (e.g., clean energy, industry) and those with a socioecological focus (e.g., hydrology, healthcare).

The eight identified research groups are: Health Care, Vegetation, Forecasting, Water, Remote Sensing, Clean Energy, Industry, and Education.

3.3. The Role of AI Technologies in Advancing the SDGs

The application of AI varies significantly across research clusters, reflecting different sectoral needs. Key roles include:

• Forecasting and System Optimization: These are the most common applications, particularly prominent in research related to SDG 7 (Affordable and Clean Energy) and SDG 15 (Life on Land), where they support resource management and operational efficiency.
• Data Mining and Remote Sensing: Widely used to extract insights from unstructured data, especially in healthcare (SDG 3) and environmental monitoring.

The most frequently applied algorithms are:

• Deep Learning and Supervised Machine Learning: These dominate applications in vegetation (SDG 15), water (SDG 6), and clean energy (SDG 7) for tasks like image classification, demand prediction, and system optimization.
• Evolutionary Algorithms: Used for complex optimization problems, such as designing efficient renewable energy layouts.
• Natural Language Processing (NLP): Emerging as a critical tool in fields with unstructured textual data, such as healthcare (SDG 3) and education (SDG 4).

4. Discussion: Gaps and Opportunities for Integrating AI and the SDGs

4.1. Critical Gaps in Social Sustainability Research

Despite the growth in publications, the application of AI remains heavily skewed towards environmental and technical SDGs. There is a profound lack of research applying AI to core social sustainability goals. Key findings include:

• SDG 1 (No Poverty): The review found no empirical studies in the most-cited literature applying AI to poverty reduction, a striking gap given its foundational importance to the 2030 Agenda.
• SDG 5 (Gender Equality): Research is minimal and primarily focuses on identifying gender bias within AI systems themselves, rather than using AI to advance gender equality.
• SDG 10 (Reduced Inequalities) and SDG 16 (Peace, Justice, and Strong Institutions): These goals are similarly underrepresented, partly due to the complexity of social data and ethical constraints.

This disparity highlights a techno-optimistic bias in the current literature, where technology is applied to quantifiable environmental problems while complex social challenges are neglected.

4.2. Disciplinary Silos and the Challenge for SDG 17

The research landscape is highly fragmented, with studies typically remaining within disciplinary silos. While AI is applied to advance specific goals like SDG 3 (Health), SDG 4 (Education), SDG 6 (Water), and SDG 7 (Clean Energy), the conceptualization of sustainability is often weak and lacks the integrated, systems-level approach required by the 2030 Agenda. This fragmentation is a major barrier to achieving SDG 17 (Partnerships for the Goals), which requires inter- and transdisciplinary pathways that the current research structure does not support.

4.3. Limitations and Future Outlook

The current body of literature demonstrates that the connection between AI and sustainable development research is nascent. The term “SDG” is often used rhetorically rather than as a framework for generating actionable, transformative insights. The focus remains on systems knowledge, with little attention to the normative and ethical dimensions central to sustainability science.

Future research must move beyond this experimental phase. To truly harness AI for the SDGs, the following is required:

1. Interdisciplinary Collaboration: Greater integration between AI experts and social scientists is needed to apply AI to complex social challenges like poverty and inequality.
2. Inclusive Data Frameworks: Methodologies must be developed to incorporate qualitative, context-rich social data.
3. Ethical Integration: Ethical considerations must be moved from the periphery to the core of AI and sustainability research to address issues of bias, equity, and justice.

5. Conclusion

Artificial intelligence holds significant potential to advance the Sustainable Development Goals, but its promise remains largely unfulfilled. The current research landscape is characterized by a narrow focus on technical and environmental goals, neglecting the critical social dimensions of sustainability. A substantial gap persists between AI applications and deep sustainability expertise, with many SDGs, particularly SDG 1 (No Poverty), SDG 5 (Gender Equality), and SDG 17 (Partnerships for the Goals), remaining severely underrepresented. To move forward, a more integrated and ethically grounded research agenda is essential. This requires fostering interdisciplinary collaboration and developing novel approaches to address the full, interconnected spectrum of the 2030 Agenda, ensuring that AI serves as a transformative tool for a just and sustainable common future.

Analysis of SDGs, Targets, and Indicators in the Article

1. Which SDGs are addressed or connected to the issues highlighted in the article?

The article analyzes the application of Artificial Intelligence (AI) across various Sustainable Development Goals. Based on the review of 792 research articles, the following SDGs are explicitly mentioned as being addressed or having significant gaps:

• SDG 1 (No Poverty): The article highlights a significant gap in research, noting that the use of AI in poverty reduction is “minimal.” It found only seven review articles and no empirical studies applying AI to this goal, despite its foundational importance.
• SDG 3 (Good Health and Well-being): This goal is frequently addressed. The analysis shows that research from countries like Italy and the United Kingdom predominantly focuses on SDG 3, with applications in health care and medical imaging.
• SDG 4 (Quality Education): This is another well-represented goal. The article states that studies focusing on SDG 4 are common among researchers in the United States, Spain, and China, exploring AI’s role in educational contexts.
• SDG 5 (Gender Equality): Similar to SDG 1, this goal is identified as “underrepresented.” The research found is primarily focused on identifying gender bias within AI algorithms and research communities, rather than broader applications for gender equality.
• SDG 6 (Clean Water and Sanitation): This goal is a major focus in the analyzed literature, particularly in studies from Iran, India, and Spain. Research concentrates on areas like watershed modeling and water resource management.
• SDG 7 (Affordable and Clean Energy): The article identifies a strong focus on this SDG, especially in research from China. AI is frequently used for clean energy optimization and forecasting.
• SDG 10 (Reduced Inequalities): The article points to a “notable gap” in AI applications for this SDG, indicating it is an area where AI is underutilized.
• SDG 15 (Life on Land): This goal is represented through research on environmental sustainability, specifically in “vegetation monitoring” and using AI to track changes in vegetation through satellite imagery.
• SDG 16 (Peace, Justice, and Strong Institutions): This is another area identified as having a “notable gap,” with limited application of AI to address its objectives.
• SDG 17 (Partnerships for the Goals): The article notes that no relevant articles were found for SDG 17 in the review. However, it emphasizes that achieving this goal is crucial and will require interdisciplinary research pathways, which AI could potentially catalyze.

2. What specific targets under those SDGs can be identified based on the article’s content?

The article points to several specific targets where AI applications are being explored or are notably absent:

1. Under SDG 1 (No Poverty):
   • Poverty Prediction: The article mentions a few examples of AI-driven “poverty prediction tools” but notes a general lack of empirical studies.
2. Under SDG 3 (Good Health and Well-being):
   • Improving Health Care: AI is used for data mining and extracting insights from unstructured data in health-care systems.
   • Advancing Medical Imaging: The article cites Italy’s focus on AI applications in medical imaging.
   • Predicting Disease Prevalence: A specific example mentioned is using AI to analyze “lifestyle predictors of dementia prevalence.”
3. Under SDG 4 (Quality Education):
   • Personalized Knowledge Building: The article suggests that generative AI will enable more nuanced investigations, including personalized learning.
   • Evaluating AI Tools in Education: Research is being conducted on specific applications like single learning platforms.
4. Under SDG 5 (Gender Equality):
   • Addressing Bias in AI: The primary focus identified is on analyzing and correcting sex and gender bias in AI algorithms, particularly in health care.
5. Under SDG 6 (Clean Water and Sanitation):
   • Sustainable Water Management: AI is applied to “watershed modelling,” “water resource management,” and “predicting water levels.”
   • Pollution Prediction: Deep learning and machine learning are employed for pollution prediction.
6. Under SDG 7 (Affordable and Clean Energy):
   • Clean Energy Optimization: AI is widely used for “system optimization” of renewable energy systems, grid performance, and layouts for wind farms and solar panels.
   • Energy Demand Prediction: Forecasting energy demand is a prominent application of AI in this sector.
7. Under SDG 15 (Life on Land):
   • Vegetation Monitoring: AI methods are effective for processing large-scale datasets for tasks like vegetation monitoring.
   • Tracking Environmental Changes: AI is used to “track vegetation changes through satellite imagery.”

3. Are there any indicators mentioned or implied in the article that can be used to measure progress towards the identified targets?

The article does not specify official SDG indicators but implies that the application and type of AI technology serve as indicators of research activity and progress in specific areas. These can be categorized as follows:

• AI Methodologies as Indicators of Progress: The prevalence of certain AI techniques indicates the focus of research.
  • Forecasting: Used to predict climate trends, water and energy demand, and disease prevalence. Its application is a measure of predictive capability in a sector.
  • System Optimization: Used to improve the efficiency of renewable energy grids and industrial processes. The degree of optimization serves as a performance indicator.
  • Data Mining and Remote Sensing: Used to translate raw data (e.g., satellite imagery, text documents) into usable insights for health care and environmental monitoring. The ability to extract actionable insights is an indicator of progress.
  • Deep Learning and Supervised Machine Learning: Their dominant use in vegetation, water, and energy indicates a focus on image classification, pollution prediction, and system optimization.
  • Evolutionary Algorithms: Their use in designing renewable energy layouts indicates progress in solving complex optimization problems.
  • Natural-Language Processing (NLP): Its emergence in health care and education indicates a growing ability to analyze unstructured textual data.
• Publication and Citation Metrics: The article itself uses these as indicators to measure the focus of the scientific community.
  • Number of Articles Published: The substantial increase in publications per year is used as an indicator of growing interest and research activity at the intersection of AI and SDGs.
  • Geographic Distribution of Research: The origin of publications is used to indicate regional research priorities (e.g., Spain’s focus on SDG 6, Italy’s on SDG 3).

Summary Table of SDGs, Targets, and Indicators

Source: nature.com