Access to Information (a2i), in partnership with the non-profit organisation GiveDirectly, hosted an international learning event today in the capital Dhaka, aimed at enhancing the delivery of humanitarian aid through innovative technological approaches. The collaboration, with the support of USAID and Google.org, alongside technical guidance from the University of California Berkeley, has made significant development over the past year in utilising machine learning and mobile data analytics to refine the targeting mechanisms for aid distribution in Bangladesh, said a press release. According to the press handout, this effort has successfully identified and facilitated cash transfers to vulnerable communities, including Rohingya refugee hosts, through mobile money solutions.

The event also served as a platform for launching a project responsive to climate crises, in collaboration with the University of Oxford and Google Flood Forecasting Initiative. This project aims to explore anticipatory action efforts globally, focusing on the effectiveness of preemptive versus post-disaster cash assistance.

The gathering brought together key stakeholders from various sectors, including government, academia, NGOs, and the private sector, to discuss advancements in social protection frameworks. The focus was on leveraging AI to enhance the targeting of aid and efficiently manage climate risks. Mollah Mizanur Rahman, a2i's Joint Project Director, emphasised the government's commitment to leveraging AI for societal betterment, governance efficiency, and national development. Additionally, Lamia Rashid, GiveDirectly's Regional Director, shared insights into their work in Togo, where machine learning and mobile data were used to provide emergency relief, inspiring the collaboration in Bangladesh.

The discussions highlighted the potential of AI-driven solutions in addressing broader social protection challenges, underscoring the importance of multi-sector collaboration in driving impactful initiatives.