Exploring how AI can boost people-centered development

With AI’s potential to transform and improve lives globally, we should consider what really works to leverage AI for people.

At the World Bank, we are learning about effective ways to deploy AI in health care, education, and jobs. Using support from the Human Capital Trust Fund, we are exploring several questions:

• How do AI tools have a real-life impact on health, learning, and jobs? Two-thirds of all pregnant women lack access to ultrasound technology that can promote healthy pregnancies. With AI-enabled portable ultrasound technology that can bring specialized prenatal diagnostics to remote and underserved communities, we are exploring its impact on pregnancy outcomes and primary healthcare systems in Ethiopia.

In many countries, more students than ever are enrolling in secondary education, but weak science, technology, engineering and mathematics (STEM) education means graduates often lack the skills needed in the job market. In Benin, we’re testing if personalized math instruction through machine learning in secondary school translates into better learning and more students that enroll and stay in STEM fields.

• How do AI-empowered public services benefit people who need them the most? We know that  jobs are critical to promote economic growth, peace, and stability, but people who live in contexts of fragility, conflict, and violence often struggle with job stability and security. In Ukraine, we are evaluating if AI-enhanced tools can provide extra help for internally displaced persons and veterans to identify their skills and get jobs.

In Poland, we're studying how to leverage AI-based digital solutions to make mental health services—from diagnosis to treatment—available for forcibly displaced populations and their host communities.

• What helps people use AI tools at scale—and what makes adopting AI more difficult? While many education systems are using AI-powered teacher coaching, lesson plans, and student assessments to support overstretched teachers, barriers to widespread adoption remain.

In Pakistan, we are exploring the constraints to more teachers using AI tools for better teaching and learning, from teacher perceptions about AI to practical issues such as lack of internet access or basic digital skills.

• Do perceptions about AI matter?

Perceptions about AI are also part of research in Rwanda. As countries shift from asking communities to identify who receives support through social protection programs to data-driven approaches enabled by AI, we’re asking if communities view algorithmic approaches as fairer and more effective.

We are exploring these questions and more through research funded by the Human Capital trust Fund in Colombia, Uzbekistan, and Peru and in multi-country studies in Egypt, Kosovo, Moldova, Tunisia, and Uzbekistan and Madagascar, Comoros, and Seychelles.

Answering these questions by using people-centered, evidence-based research can help roll out AI-empowered systems in health, education, and jobs to boost development more effectively and efficiently.

Now more than ever, we must explore AI’s potential to transform and improve lives globally. Watch this space for what we learn from this research and more.