Skip to main content
 

Fostering responsible innovation

IDRC is at the forefront of global development efforts, championing research for development that empowers communities, accelerates progress and builds sustainable futures.

IDRC investments across the Global South in 2023-24

 

Artificial intelligence (AI) is rapidly transforming the global landscape. From revolutionizing health care to tackling climate change, AI holds immense potential to improve lives around the world. As the global conversation evolves, discussions are swirling about the ethical implications of developing and deploying AI. Key questions are being raised about potential biases in algorithms, the impact that AI will have on jobs and how to ensure that AI will benefit all of humanity. These discussions are crucial for shaping a future where AI serves as a force for good, promoting sustainable development and a more equitable world.

This year, IDRC’s annual report spotlights the power of research to support the development of responsible AI innovations. IDRC began to invest in high-quality AI research and innovation in the Global South nearly a decade ago. Since then, our investments have grown into a vibrant AI research ecosystem — a complex and interconnected network of entities, organizations and resources that contribute to the creation, dissemination and application of knowledge.

Today, IDRC is a world-leading supporter of responsible AI research and innovation to advance international development goals. Our research investments across Africa, Asia, Latin America and the Middle East are enhancing the capacity of individuals and institutions to bring about a brighter, more resilient and sustainable future. 

Guided by our Southern partners, IDRC’s efforts to develop AI research and innovations are deeply rooted in local needs and aspirations. We believe that people facing development challenges are the best placed to identify issues and find solutions. That is why we support academic institutions, local governments, civil society organizations and grassroots communities to co-create AI innovations that are contextually relevant and culturally sensitive. 

In the right hands and with the right focus, AI applications can be used to change lives for the better, and, as our investments show, we recognize their immense potential to augment health-care capacity, mitigate the effects of climate change, revolutionize education models and so much more.

However, we are also mindful of the significant risks that AI can bring. In the wrong hands, AI has the power to exacerbate existing inequalities and threaten jobs, civil liberties and human rights. This is where responsible AI research becomes crucial. By supporting the research and development of evidence-based, responsible AI systems that are safe, inclusive, rights-based and sustainable, we can identify and address biases in the technology while also gaining insights into the benefits and potential uses of responsible AI.

As a dynamic and urgent global conversation about AI gains momentum, IDRC’s research investments have proven to be prescient. Canada is a country actively involved in conversations about AI, and the well-established research supported by IDRC is positioned to make significant contributions. 

In this report, we delve into IDRC’s efforts to build sustainable AI research capacity in the Global South — and support the many innovations that result — while concurrently addressing the risks associated with AI in the development sphere. Throughout, we explore the diverse applications of AI innovations across our programming areas while anchoring our endeavours in the principles outlined in IDRC’s Strategy 2030.

Highlights
Media
A group of anxious villagers in Sariakandi, Bangladesh, stand beside the River Jamuna, where erosion is eating into the river banks.
Panos/GMB Akash

Inclusive solutions through responsible AI: IDRC’s growing impact

After years of partnering with emerging AI research and innovation leaders in the Global South, IDRC and the Swedish International Development Cooperation Agency launched the AI for Development (AI4D) Africa program in 2021. The four-year, $20-million partnership saw the creation of an AI policy research network, two AI policy think-and-do tanks, two AI laboratories and a scholarship program at public universities, and four AI innovation research networks. These have resulted in nearly 100 locally led innovations — from improving climate resilience to supporting a feminist approach to AI research and development.

This partnership will benefit all countries and ensure that developing countries are not left behind in the AI revolution.

 

This year, IDRC partnered with FCDO on an expanded, $70-million phase of the AI4D program that will run until 2029. Called the AI4D: Responsible AI, Empowering People program, it will build on the foundations of AI4D’s first phase by coordinating IDRC’s AI programming in Africa with work we support in Asia, Latin America and the Middle East — work such as the Global Index on Responsible AI, the Feminist AI Network and the AI for Global Health initiative — to create the greatest possible impact.

Unlocking equality, combatting AI bias

The Feminist AI Research Network (FAIR), part of the new phase of AI4D at IDRC, is a global network of scientists, economists and activists finding ways to make AI and related technologies more effective, inclusive and transformational. One key threat embedded in the emergence of AI is its potential for deepening inequalities, especially for women and vulnerable populations. FAIR supports projects that identify and correct digital biases and that develop AI solutions reflecting feminist principles.

Media
Women transit user in Manila
IDRC / David Hogsholt, Panos Pictures

Harnessing AI for economic policymaking in Africa

As AI continues to transform various sectors, its applications in government services and policy development remain largely unexplored. The African Center for Economic Transformation is researching the potential of AI to revolutionize economic policymaking in Africa as part of IDRC’s AI4D program. This project aims to: 

  • generate evidence on how AI can be used to improve economic policy decisions in Africa, leading to faster and more inclusive economic growth 
  • research and experiment in collaboration with policymakers, with the goal of creating a realistic understanding of AI’s potential in the policy sphere 
  • build capacity by bringing together a community of policymakers, practitioners and researchers who share knowledge 
  • promote equality by recognizing the importance of addressing potential biases in algorithms and data, and ensuring AI tools are implemented in a way that benefits all citizens.
     

Empowering health systems for equity and inclusion

The use of AI is transforming how health systems and services are planned and delivered globally. AI innovations that prioritize human needs and uphold fundamental rights have the potential to streamline health-care delivery and accelerate progress toward the health-focused Sustainable Development Goals. Such innovations can be integrated across various aspects of health-care systems, from largescale public health monitoring to onsite services at the point of care. 

However, evidence is needed in low- and middle-income countries on how AI can be used to strengthen health systems, improve health equity, support gender equality and inclusion, and ensure a local supply of skilled people, credible data and computing infrastructure. 

The AI for Global Health (AI4GH) initiative, also part of IDRC’s new AI4D programming, is a five-year $15.5-million IDRC investment that funds interdisciplinary research projects across Africa, Asia, Latin America and the Caribbean and the Middle East. It brings together experts from various fields, including health care, data science and AI, to collaborate on developing responsible, locally driven AI solutions that support better and more equitable preparedness for and responses to epidemics and pandemics and improve sexual, reproductive and maternal health for women and girls.
 

Preparing for the next outbreak

Disease outbreaks are increasing in frequency and severity, and climate change is worsening the world’s health and social inequities by creating climate “hotspots” where diseases are more likely to emerge. The Global South AI for Pandemic Preparedness and Response (AI4PEP) Network is supporting 16 hubs composed of interdisciplinary researchers and policymakers across Africa, Asia, Latin America and the Caribbean, and the Middle East to detect, respond to and control new disease outbreaks before they become epidemics or pandemics. 

This global initiative aims to smash siloes. It brings data scientists, epidemiologists, physicists, mathematicians and software engineers together with disaster and emergency management, clinical public health, citizen science and community engagement experts. Interdisciplinary teams of researchers work to improve early detection, early-warning systems, early response, and ways to mitigate and control developing epidemics with AI as the entry point. 

 

AutoAI-Pandemics stands as a potentially transformative force, poised to address critical societal challenges by offering solutions to misinformation and by providing efficient computational tools… (It) aims to significantly reduce the impacts of epidemics and pandemics, either avoiding or reducing future global crises like that experienced in early 2020.

 

As just one example of an AI4PEP project, a group of researchers at the University of São Paulo in Brazil is developing AutoAI-Pandemics, a platform designed to make AI accessible to professionals — such as biologists, physicians and epidemiologists — who are not experts in data science or machine learning. 

The researchers aim to accomplish several things: 

  • automatically analyze disease outbreaks and find the best ways to stop them 
  • discover new drugs and understand the genetic makeup of pathogens 
  • fight misinformation about diseases by giving people reliable information 

The AutoAI-Pandemics research team wants their platform to be easy to use. They will share their findings in articles and train doctoral students about how to use AI in health care.

Media
Dr. Oluwafemi Ogunrayi conducts a check-up during an antenatal visit at Okpekpe Community Primary Health Centre in Edo State, Nigeria.
IDRC/ANDREW ESIEBO

An AI breakthrough for pregnant and postpartum women 

In remote areas of Kenya, where access to high-quality maternal health-care services can be scarce, a quiet crisis has loomed. In 2018 alone, 38,000 mothers and newborns lost their lives because they lacked access to reliable maternal and newborn care. 

Non-profit organization Jacaranda Health has developed a game-changing tool for maternal and newborn care. Called PROMPTS, the tool is a system based on text messages (SMS) that lets pregnant and postpartum women get assistance from a human helpdesk, no matter where they live. 

But PROMPTS has 350,000 users and its success brought challenges. With 1.1 million SMS messages flooding in each month, the helpdesk at Jacaranda Health needed a triage tool to help them deliver timely interventions for women in critical need. Jacaranda Health worked with a team of researchers to develop Triage for Mothers using AI (TRIM-AI), a state-of-the-art natural-language processing framework. 

TRIM-AI assesses the emergency level of pregnant mothers based on the content of their SMS messages. In a comparison, the TRIM-AI version of PROMPTS was 17% better at spotting high-risk conditions in texts from Kenyan mothers. And it reduced the helpdesk’s workload by 12%, giving them more time to focus on helping mothers and babies.

Amplifying Southern voices for ethical AI development

While the current discourse on AI governance is lively and productive, it lacks sufficient representation from Southern voices and perspectives. To mitigate harm and combat biases effectively, it is crucial to involve experts from Africa, Asia and Latin America in shaping recommendations and norms. 

IDRC is funding Southern partners to develop guidelines, frameworks and regulatory mechanisms that promote the responsible and ethical deployment of AI in development contexts. We facilitate multi-stakeholder dialogues, knowledge exchanges and partnerships to develop standards and best practices for responsible AI deployment that fosters human rights, privacy, accountability and social inclusion. The vibrant networks that we fund work towards common goals, fostering a future where AI serves as a force for positive change in development.

Simply adopting new or innovative approaches to old problems is insufficient. We need to proceed with care and caution, including by marshaling insights from the vast existing knowledge about the promises and pitfalls of leveraging technology and data science for development.

Luciá Mesa Vélez, the Ladysmith Collective, an IDRC-supported partner

Advancing global governance for responsible AI 

IDRC and Global Affairs Canada are spearheading an initiative that will empower governments, civil society organizations and others with evidence to advance the principles of rights-based AI governance. While strides have been made in delineating ethical frameworks for AI, the need now lies in translating these principles into actionable strategies across public and private sectors. 

Enter the Global Index on Responsible AI. Led by the South Africa-based think tank Research ICT Africa, this project aims to deliver a comprehensive and dependable assessment of AI development and use across more than 120 countries. Its goal is to better inform international cooperation and ignite local action for responsible AI. 

The project team has established a robust measurement framework of 87 indicators and 29 thematic areas designed to gauge the responsible deployment of AI. Harnessing the expertise of 11 research institutions, predominantly hailing from the Global South, and supported by 135 country experts, the project scrutinizes the landscape of responsible AI practices against the measurement framework. The index was released in June 2024.

Media
 Image showing computer graphics
iStock

Aligning AI with human rights and democratic values

The Global Partnership on Artificial Intelligence (GPAI) brings together experts from the public, private and not-for-profit sectors worldwide to foster the responsible development and use of AI in alignment with human rights and democratic values.

Through its working groups, GPAI focuses on key areas such as responsible AI, pandemic response, data governance and innovation. In 2020, GPAI established the International Center of Expertise in Montréal for the Advancement of Artificial Intelligence (CEIMIA) to lead initiatives in responsible AI, pandemic response and data governance.

To enhance GPAI's inclusivity and perspective from the Global South, IDRC is funding an AI specialist from the region. Thomas Hervé Mboa Nkoudou, a Cameroonian researcher now based in Montréal, is collaborating with CEIMIA, the working groups and the Canadian AI community to integrate Southern countries' needs and interests into GPAI's overall efforts toward responsible AI adoption. His role is pivotal in fostering stronger ties between GPAI, CEIMIA and AI stakeholders in the Global South.
 

Combatting misinformation 

In recent years, misinformation has spread rampantly across the global online media landscape, creating what experts call an "infodemic." The problem is especially urgent in the Middle East and North Africa, where a lack of independent media has led people to rely heavily on social media for news. 

IDRC is funding research to assess the current state of misinformation in Arabic by gathering data from Arabic factchecking initiatives and creating a unique multilateral database. Using natural language processing techniques, the research team, led by Wael Alalwani of Open Transformation Lab Inc., will analyze data to identify trends in how misinformation impacts minority groups and the public perception of misinformation and how it spreads.
 

Addressing data disparities 

Data is the foundation of AI innovations and serves as the foundation for algorithms to learn and evolve. When data does not exist, is outdated or does not represent underserved populations, the result can be a biased, inaccurate AI tool.

In 2020, IDRC, the Rockefeller Foundation and other partners united to create the Lacunda Fund, the world's first collaborative effort to fill critical AI data gaps in agriculture, language, health and climate. To date, 16 projects have received funding: 10 for language and six for agriculture. Researchers have developed 13 datasets, all of which are openly accessible. 

Agriculture awardees have produced training datasets across sub-Saharan Africa to support various agricultural needs. Language awardees have produced text and speech datasets for natural language processing technologies in East, West and Southern Africa. 

Every language and culture has a story to tell, and one’s native language speaks to one’s soul…In this project, we will get the chance to work with selected native speakers across Kenya, involve students in data collection and annotation, and mentor them into building (natural language processing) tools for African languages.

KenCorpus project team

In the KenCorpus project, a team of researchers at Maseno University in Kenya has developed rich textual and speech data resources for the following languages spoken in Kenya: Kiswahili, Luhya (including three dialects) and Dholuo. The project is addressing the underrepresentation of African languages in the realm of AI and natural language processing. Among its many potential benefits: helping preserve the languages it represents and serving as a foundation for language-learning materials, such as textbooks and language-learning apps. 

The dataset has been downloaded an incredible 250,000 times!

Media
Abdul Karim Mbatch watering some of thousands of coconut trees planted on Banjul’s beach to help stem coastal erosion from climate change.
Sylvain Cherkaoui/Panos Pictures

Looking ahead

The dual disruptions of AI and climate change emphasize the urgency for inclusive and sustainable development strategies that mitigate risks, build resilience and empower individuals and communities — particularly women and the vulnerable — to thrive. As the world increasingly turns its attention to the transformative potential of AI innovations, IDRC will continue to lead by fostering innovative solutions through research. 

As one example, IDRC announced in November 2023 the launch of FutureWORKS, a five-year, interdisciplinary, Southern-led research initiative that addresses the future of work in the Global South amid these evolving realities. Five regional hubs, each of which will receive $1.7 million, have been selected to develop and lead research ecosystems in Asia, Africa, the Middle East and Latin America that reflect local talent, needs and interests. 

IDRC is proud to collaborate with Canada as we navigate the evolving landscape of technological advancements. Leveraging Canada's leadership in AI policy, we will continue to support the development of inclusive and ethical AI policies, practices and innovations that address the unique challenges and local realities of the Global South. 

As AI increasingly shapes economies and societies worldwide, we will work to ensure that emerging technologies benefit all — especially vulnerable communities — and promote sustainable development and equitable access to opportunities. Through our continued efforts, IDRC aims to catalyze positive change, driving progress toward a more inclusive and prosperous future for all.

Governance

Governance

The Board of Governors is responsible for the Centre’s stewardship: it sets strategic direction and oversees operations. The Board acts and conducts all of its business in accordance with the IDRC Act, IDRC general by-laws and governance best practices. The roles and responsibilities, composition and organization of the Board are described in detail in its Charter.

Go to Governance 
 

Transparency

IDRC is accountable to Parliament and all Canadians for its use of public resources. Here are some of the measures in place that help us meet or exceed the standards set by the Government of Canada for accountability and transparency.

Go to Transparency

Finance

Management’s Discussion and Analysis and Financial statements

This Management’s Discussion and Analysis (MD&A) provides a narrative discussion of the financial results and operational changes for the financial year ended 31 March 2024. This discussion should be read alongside the Financial statements starting on page 35, which were prepared in accordance with the International Financial Reporting Standards. All monetary amounts are presented in Canadian dollars unless otherwise specified.

Read the full Annual Report 2023–2024

Download

Read previous annual reports