AI technologies can be one of several tools used to address local community challenges, if developed and deployed by those communities. Cultural sensitivity and familiarity with the day-to-day challenges are key to creating AI systems that are fit-for-purpose and overcoming barriers to adoption.
On April 17, 2025, fellows participated in a discussion on the topic of using AI to address community challenges in the context of countries in the Global South. To seed the conversation, four fellows presented case studies and research findings that highlighted existing mechanisms for ensuring AI systems meet the needs of the communities for which they are developed and deployed.
Discussion Summary
The fifth and final discussion session for the second cohort of the Global Perspectives Responsible AI fellowship program focused on how AI is being used to solve community challenges. AI innovation continues to grow alongside increased internet and mobile penetration rates across the Global South, leaving governments and communities to navigate both simultaneously and evaluate whether and how AI can be used to bridge technological divides. The most efficient way to navigate these twin digital transformations has been to engage in multistakeholder discussions around the development and deployment of AI systems and models. When built with only government in mind, new technologies may not work well for communities, and when built at a hyper-local level, AI technologies may not be sustainable. Fellows therefore argue for a feedback loop between governments, private sector, academia, and community members to collaboratively solve problems ranging from artistic expression to real-time flood monitoring.
Turkey
To fully harness AI as a tool for cultural transformation, access to AI technology needs to be democratized and built in ways that empower artists. For example, the intersection of AI and intellectual property rights has long been debated and re-emerged most recently with the use of generative AI models to ‘ghiblify’, or to transform an image into the anime style of Japan’s Studio Ghibli. While there is much work to be done in to create effective regulatory measures that are best suited for the AI-age, overcoming barriers to AI adoption by the artistic community will require respecting artists and their intellectual. Moreover, for AI to be used by artists across the Global South, models need to be trained in specific cultural realities.
Beyond the art space, generative AI can be used to enhance creativity, efficiency, and problem-solving across a range of diverse projects. Turkey’s AI ecosystem—nurtured by the Turkish Artificial Intelligence Initiative (TRAI)—invests widely across the industry to incorporate AI into many startups and private sector companies in the country. By prioritizing Turkish culture within the growing AI ecosystem, TRAI works to promote a deeper understanding of AI models across Turkey and make their systems more accessible not only domestically, but globally.
India
India faces a unique challenge when it comes to AI skilling and technological literacy. Though India has been poised as a pioneer of AI innovation in the Global South since 2018 and aspires to build indigenous LLMs, it has lately seen a stark decline in the quality of technical education. It is estimated that only 25 percent of engineering graduates are employable in India. This can be attributed to the existing digital divide and limited access to AI education in rural areas, a lack of preparedness to enter the AI workforce, the exclusion of marginalized communities from AI development, and a lack of responsible AI education and support programs. One solution has been the Responsible Computing Challenge (RCC), a community-driven and youth-led initiative that engages with future technologists early on. The RCC is participatory by design and educates young technologists on a variety of issues from diversity and fairness to environmental awareness. As the RCC grows, it aspires to lead faculty training initiatives, a student ambassador program, and workforce development centers that will work with and empower tribal groups and marginalized communities to solve community challenges and preserve culture using AI tools.
India also continues to make progress in advancing equity for disabled communities. Marian College – Kuttikkanam developed a program on Equitable Digital Access for undergraduate students and an augmented reality handbook on equity in digital accessibility which can be reviewed through immersive technologies. Students taking the Equitable Digital Access course at Marian College – Kuttikkanam developed WebEase 1.0, an open-source web accessibility widget enhancing inclusivity across platforms. Moreover, students at DJSV College participating in the RCC developed PhysioPlay, an AI-powered diagnostic training tool for aspiring physiotherapists.
Indonesia
In Indonesia, AI has not only been proven as a beneficial community management tool but has also empowered communities to act. PetaBencana.id is an AI system that uses both crowdsourced data and IoT sensors to detect and mitigate environmental disasters. The application proved to be invaluable in 2017 during the Jakarta floods, enhancing community resilience and government response times. Public universities have also begun integrating AI technology in the form of teaching assistants. The AI assistants are trained to answer students’ questions and instruct on over 1,000 online courses, designed to democratize access to quality education. The assistants, developed by Universitas Terbuka, have reached 60,000 students across the country, bridging the urban-rural education gap. Korika, an organization founded through Indonesia’s 2020 National Strategy for AI, also works to close the gap, targeting rural and marginalized communities. Korika facilitates mentorship opportunities through its training program and volunteers from multinational companies and local tech communities.
Other responsible AI efforts in Indonesia cover cultural sensitivity, data sourcing, and more. Given the cultural complexity of Indonesia (700+ languages and 1,340+ ethnic groups), there is no one cultural standard. Nonetheless, the IndoCulture Dataset aims to integrate cultural commonsense reasoning into Indonesian AI technologies and reduce bias. In 2023, the Ministry of Communication and Information (KOMINFO) issued an AI Ethics Guideline which led to a 2024 collaboration between KOMINFO, Korika, and UNESCO to pilot a new AI Readiness Assessment (AI-RAM). Korika published their own guide for developing AI ethically based on Pancasila – a locally held belief in God, humanity, unity, democracy and social justice. The guide incorporates technical safety, stakeholder collaboration, and human-centric practices.
Kenya
The AI ecosystem in Kenya supported the development of home-grown AI-powered solutions such as Twiggy, Ushahidi, M-Schule, and Zuri. Twiggy is an AI chatbot via WhatsApp developed and deployed by Infobip and Twiga (a tech company that focuses on connecting consumers and suppliers with high quality products in Africa). With Twiggy, vendors can better scale onboarding initiatives for small and medium-sized vendors whose workers may have low digital access. The Ushahidi Platform uses citizen-generated, crowdsourced data to better geolocate and classify incident reports in crisis scenarios, leading to more community engagement and faster election monitoring and disaster relief. M-Schule is an educational SMS platform that delivers personalized learning to students in both English and Kiswahili, matching the students’ pace and understanding to empower learning in informal settings. Zuri is the AI chatbot for Safaricom and is available across platforms 24/7 to reduce the need for human customer service agents while offering immediate assistance.
To further advance its position as a regional leader in AI adoption, Kenya released its National AI Strategy 2025-2030. The strategy has six main components, each with their own set of objectives and initiatives: AI Digital Infrastructure; Data Governance and Sovereignty; AI Research and Innovation; Talent Development; Governance and Legal Frameworks; and Ethics, Equity, and Inclusion.
Conclusion
AI is being used to solve community challenges across the Global South from access to the arts to disaster response. AI-based solutions thrive in ecosystems where governments, private sector, academia, and community members engage in a meaningful exchange of information and resources. In these ecosystems, solutions can be built informed by both top-down and bottom-up priorities, meaning that the government can develop and deploy tools that benefit their communities, and those communities can influence how those tools work and/or develop their own. In any case, the government is an important enabler for AI ecosystems, and when governments prioritize responsible AI development, invest in developing and deploying national strategies for AI, and devise effective regulatory frameworks that protect their population and enable innovation, local talent is better suited to create systems that work for communities and narrow the growing AI divide. Equally important is the influx of participatory design practices that include marginalized and disconnected communities in the development process, so that future AI systems and models better understand cultural and linguistic nuances.
This summary captures the extensive discussion and insights shared during the meeting held on April 17, 2025, at 10:00am EST while maintaining the anonymity of individual participants as per the Chatham House Rule.
A special thank you to Isaac Halaszi, Research Assistant for the Strategic Foresight Hub at the Stimson Center, who contributed to this event summary.