For many people supported by the humanitarian sector, 2024 was a bad year. of Latest UN estimates A record 120 million people have been forced to flee violence and disaster, a figure that has doubled in the past decade. The broad figure for people in need of humanitarian assistance is 300 million and rising due to escalating conflicts and the growing impact of the climate crisis. Progress has also been made in achieving the United Nations Sustainable Development Goals. stagnant or decreasing in more than half of the vulnerable countries. Children born in those countries are 10 times more likely to fall into poverty than children born in stable conditions.
These unprecedented numbers demonstrate the need for new humanitarian aid, powered by digital and AI-powered technology. We have been (rightly) debating the risks and benefits of AI for years, waiting for the promise of “AI for good” to arrive. In 2025, that moment may finally be on the horizon across the aid, development and humanitarian sector.
When used properly, AI can open new frontiers for humanitarian work in terms of scale, speed, scope, personalization, and cost reduction. In my organization, international rescue committee (IRC), and an in-house research and innovation lab; airbell, Exploring the application of AI in humanitarian programs. Solutions are emerging in three key areas: information, education, and climate change, each powered by promising public-private partnerships and collaborations.
For refugees forced to flee conflict, for example, the priority is providing timely, accurate and context-specific information about who to trust and where to find services and safety. . World information project “Signpost” With support from Google.org, Google’s philanthropic arm, we partnered with IRC, Cisco Foundation, Zendesk, and Tech for Refugees to provide critical information to millions of displaced people through digital channels and social media, and to eliminate misinformation. It disempowers smugglers who thrive on false information and false information. It can also save lives along migration routes. As this work progresses, Signpost is establishing an “AI Prototyping Lab” to reduce risk and evaluate the effectiveness of generative AI across the humanitarian sector.
Humanitarians are also exploring the potential of generative AI to enhance and personalize the education of the world’s 224 million crisis-affected children. Big challenges include testing and enhancing ChatGPT’s potential in local languages. For example, AI models cannot understand African languages. Relapa AIThe African AI Research and Products Institute is working to change this. develop a new language How OpenAI brings AI to Africa We have started offering Low-cost access to ChatGPT for nonprofits.
OpenAI also supports the development of: Append AIis a global AI-driven educational chatbot platform that provides crisis-affected children, teachers, and parents with personalized digital learning experiences at scale via a messaging platform, while also providing local language support to children, teachers, and parents affected by the crisis. We are testing and enhancing the potential of ChatGPT.
Finally, the power of artificial intelligence is being expanded to protect communities facing the severe impacts of extreme weather events. In cooperation with NGOs, governments and the United Nations, Google launched The AI-powered “flood hub” can now predict floods in 80 countries. Collaborating with Google.org, IRC and NGOs give directlyILeverage machine learning It aims to establish a forecasting system that triggers early warning and cash transfers ahead of catastrophic climate disasters in northeastern Nigeria.
Israeli scholar and historian Yuval Noah Harari explained about artificial intelligence This is the most dangerous technology we have ever created, and also potentially the most beneficial. By 2025, these benefits will reach the world’s poorest people.
