AI in low-income countries: youth optimistic about education and work opportunities, new Restless Development survey finds

Overview

The Conversation presents key findings from the 2026 Youth AI Survey conducted in ten lower‑income countries in Africa and South Asia, with the UK included for comparison. The survey of 1,864 young adults aged 18‑35 reveals a generally positive outlook on AI's potential to improve education and work prospects, even as access barriers persist.

• Lower‑income country youth expect AI to boost education and learning opportunities, with many citing personal digital skills development as a primary benefit.
• About half of respondents foresee AI improving employment opportunities within five years, though concerns about future income remain (32‑50% per country).
• Positivity is strongest in Tanzania and Zambia, with women in these countries notably upbeat; in the UK, some young people are more cautious about AI’s impact on work.
• Common worries center on data ownership, corporate power, language marginalisation, and environmental impact, highlighting the need for inclusive governance and safer AI deployment.

Author and publisher: The Conversation, based on the Youth AI Survey 2026 conducted by Restless Development and researchers from Makerere University and Cambridge University.

Introduction

The Conversation reports on the 2026 Youth AI Survey, a joint effort by the global charity Restless Development and academics from Makerere University in Uganda and Cambridge in the United Kingdom. The survey spans ten countries in Africa and South Asia (Angola, Kenya, India, Nepal, Nigeria, Sierra Leone, Tanzania, Uganda, Zambia, Zimbabwe) with the UK included for comparison. It gathers responses from 1,864 young adults aged 18‑35 and reveals a generally positive sentiment about AI’s potential to enhance education and social life, alongside pragmatic awareness of access barriers and risks associated with AI use.

Survey scope and usage patterns

Two‑thirds of respondents report using AI daily, with Meta AI emerging as the most common tool, followed by Claude. The extent of AI adoption is notable given obstacles such as high data costs, weak internet connectivity, and limited digital literacy—issues more acute in most surveyed African countries than in the UK, where ethical concerns tend to limit usage. The UK participants demonstrate a different pattern, with governance concerns playing a larger role in their AI usage decisions rather than purely access barriers.

Respondents describe AI primarily as a means to learn new skills and to aid in job searches. Across most lower‑income countries, attitudes toward AI’s role in work life are broadly positive, though a sizable minority expresses worry about potential negative impacts on future earnings (32%–50% within each country).

AI work opportunities and economic expectations

Direct income from AI tools remains uncommon (only about 14% currently earn income from AI tools). Yet roughly half of those surveyed anticipate AI improving their employment and economic opportunities within five years. This aligns with expectations that AI could accelerate access to learning, upskill, and facilitate more efficient job searching. In practice, many respondents already use AI to learn new skills and to search for jobs, signaling early adoption of AI as a career enabler in lower‑income settings.

Biggest AI concerns and cultural considerations

Despite the optimistic tone, several concerns persist. A notable portion of respondents worry about data ownership and the power wielded by AI companies. Some participants describe AI as “Americanised,” arguing that AI systems are built around Western norms and may fail to capture local linguistic and cultural nuances. Language marginalisation is highlighted by examples such as Shona and Luganda, where AI spellings, grammar, and interpretation can fall short of local realities, potentially marginalising Indigenous languages further.

Other concerns track broader implications: environmental impact, mental health risks, privacy, and governance. The sentiment analysis indicates higher usage of AI correlates with higher levels of concern in several domains, including data ownership and governance. Respondents also note that cheaper and more reliable data access, improved connectivity, and culturally aligned AI designs are essential for safer and more inclusive AI adoption.

Gender differences in AI knowledge and use

There are notable gender differences. Significantly more young men than women report having “a lot” of AI knowledge (38% vs 23%), and men are more likely to use AI to search for work (33% vs 17%). These gaps are likely rooted in disparities in data access, digital literacy, and broader social roles, including care responsibilities that disproportionately affect women. However, in some contexts, such as urban Zimbabwe, respondents express strong concerns about data ownership, illustrating that the data‑ownership issue affects both genders though differently across regions.

What young people want from AI

Participants stress a demand for cheaper, more reliable data and internet connectivity, tools that fit local lives and languages, and AI systems that respect local cultures. Several initiatives aiming to improve AI inclusivity in lower‑income countries are discussed, including frugal AI approaches and regional language projects like Masakhane for African languages and QuechuaBase in Peru.

AI safety, governance, and policy implications

Most participants frame AI safety as a governance challenge rather than a purely technical one. They support governance and regulation led by governments, with global policy discussions informing national AI frameworks. The survey’s findings echo broader concerns raised by international bodies that effective AI governance is essential to ensure responsible deployment, protect privacy, and avoid cultural erosion and language marginalisation.

Conclusion and takeaways

The Youth AI Survey 2026 highlights a nuanced picture: AI is seen as a significant lever to improve education and work prospects in lower‑income countries, but realizing this potential requires addressing data affordability, language inclusion, and governance. Initiatives that adapt AI to local languages and contexts, alongside policies that protect data rights and privacy, are critical to shaping AI as a force for inclusive development rather than a driver of further inequality.

Key takeaways

• Optimism about AI’s role in education and employment is widespread among young people in Africa and South Asia, even where access barriers exist.
• Usage is increasing, with two‑thirds reporting daily AI use and AI being used mainly for skill development and job searching.
• Data ownership, company power, and language inclusivity are major concerns that must be addressed through governance and inclusive AI design.
• Calls for frugal AI, local language initiatives, and government‑driven regulation underline a need for safe, culturally aware AI deployment.