| Resumen: |
The use of AI systems in social protection is a reality, not a future prospect. With applications in at least 48 countries as of March 2026, this number is likely to grow. As the consequences of poor AI governance fall on the most vulnerable populations, it is imperative that the responsible use of AI in social protection be documented so that it can inform governments, policymakers and social protection practitioners moving into this space. To chart a way forward that ensures responsible, innovative and sovereign AI adoption we must learn from the past to guide future decisions. This Global Evidence Review examines what AI in social protection looks like in practice: where it exists, in what form, and with what evidence. |
