Artificial Intelligence in the Development Sector: What’s It All About?

CanWaCH members have recently expressed interest in exploring the use of artificial intelligence (AI) in the development sector and in learning more about ways to harness its power to achieve the Sustainable Development Goals.

When used responsibly, AI can transform the way international development and humanitarian aid organizations operate and make decisions. AI tools can help: 

• Improve work efficiency, increase capabilities, reduce workloads, and enable staff to prioritize strategic tasks.
• Gather, organize, and analyze massive datasets; enhance data quality and reduce errors; identify trends and patterns; synthesize reports; and offer real-time insights, 
• Streamline monitoring and evaluation (M&E) processes.
• With administrative and financial tasks such as writing requests for proposals.
• Organizations learn from past projects and incorporate lessons learned into new initiatives. 

Leveraging AI’s Potential to Empower Practitioners and Tackle Global Challenges 

At a time when our sector is forced to deal with a multitude of crises and when resources and funding are limited, using new and innovative tools such as AI can help to improve our capabilities. AI’s potential to support organizations in delivering their mission and maximizing their impact extends to many areas, such as poverty alleviation, agriculture, food and water systems improvement, crisis response, and climate action. 

For instance, in collaboration with Google AI, the World Food Programme developed the open-source tool SKAI. The tool uses satellite imagery and machine learning algorithms to provide real-time information post-disaster, thereby empowering organizations to make effective data-driven decisions with unprecedented precision and speed. In the area of climate action,  a partnership between UNFCCC with Microsoft is leading to the use of AI and advanced data technology to track global carbon emissions and assess progress under the Paris Agreement.

Harnessing AI for Health

AI could also transform the global health sector by changing how health systems are planned and how services are delivered. AI tools can be used to enhance health outcomes by “improving medical diagnosis, treatment, self-care, and person-centered care, strengthening clinical trials, and supplementing healthcare professionals’ knowledge, skills and competencies,” as the World Health Organization (WHO) analyses. 

AI can organize and leverage often siloed and disorganized health data to provide insights into health trends, disease patterns and evolution and help predict health risks. 

The International Development Research Centre (IDRC) is one of the Canadian organizations that has been investing in applied research across a number of domains to advance the public good with the use of AI for development. IDRC’s Artificial Intelligence for Global Health (AI4GH) initiative funds researchers working on contextualized and responsible AI solutions to healthcare challenges in low- and middle-income countries. AI4GH has established four regional hubs, some focusing on catalyzing AI to advance maternal, newborn, and child health (MNCH) and sexual and reproductive health (SRH). For example, the Africa hub‘s ten sub-grantees are implementing AI projects in Nigeria, Ethiopia, Namibia, Tanzania, Ghana, Kenya and Uganda on four priority themes: maternal health, STIs, adolescent sexual reproductive health and HIV.

Over the past few years, several organizations have been harnessing AI’s potential to enhance the delivery of MNCH and SRH services and to provide essential information in these areas. UNFPA’s ‘Just Ask!’ is one example of an AI digital engagement platform for adolescents and young adults aiming to help them learn about their sexual and reproductive health and rights. 

Recognizing Risks 

While AI brings a wave of possibilities for the development sector, the potential risks and harms cannot be ignored. One of these major risks is that AI can amplify societal biases, including gender biases, and reproduce stereotypes, since AI systems may replicate and exacerbate the biases and assumptions embedded in their design or the datasets they’re trained on. 

Other concerns include the compromise of privacy and the speed at which AI technologies are being deployed, sometimes without a full understanding of performance and risks. Taking a closer look at the Global South, particularly in countries with limited institutional capacity and legal safeguards, the impact of these risks could be exacerbated. Recognizing these risks, the WHO recently released a new publication listing key regulatory considerations on AI for health, including the importance of transparency and documentation, fostering collaboration among all stakeholders from developers to health workers, ensuring privacy and data protection, and risk management.

Want to learn more about ways to harness the power of AI in your work? CanWaCH is organizing an interactive training series from January 24, 2024! Join us virtually for: “Chat GPT Unleashed: Ignite Your Monitoring and Evaluation Journey with AI.” The three-part series will give you the tools to navigate the challenges of gender-transformative global health programming with AI as your ally. Learn more and register here. Limited to 40 spots!