Artificial intelligence (AI) is reshaping how businesses operate, compete and grow. For entrepreneurs, start-ups and micro, small and medium-sized enterprises (MSMEs) in developing countries, AI offers a chance to raise productivity, innovate, cut costs and reach new markets. But without the right conditions, it could also deepen existing divides.
This report examines what it takes for entrepreneurs in developing countries to adopt AI in practice – and what holds them back. Drawing on global evidence, surveys from UNCTAD’s Empretec network and case studies, it shows that successful AI adoption depends less on technology itself than on the surrounding ecosystem – skills, data, finance, regulation and trust.
It highlights the central role of learning and skills development in helping entrepreneurs move from awareness to effective, responsible use of AI in their day-to-day business decisions.
Key takeaways
- AI can boost productivity, but gains are uneven
AI is already being used by MSMEs and start-ups across core business functions, from marketing and customer service to logistics, finance and product design. Simple tools such as chatbots and other off-the-shelf applications are easier to deploy and often deliver quick returns. Large language models, in particular, are emerging as foundational building blocks that allow smaller firms to adopt AI quickly and at relatively low cost.
But more advanced applications require stronger digital capabilities and skills, which many smaller firms lack. As a result, adoption and outcomes vary widely across countries and firm sizes.
- Connectivity remains the first hurdle
AI relies on internet access, data and computing power. Yet only 27% of people in low-income countries are online, compared with over 90% in high-income economies. In least developed countries, about 65% of the population remains disconnected, keeping AI out of reach for many entrepreneurs.
- Skills and managerial gaps slow adoption
Many entrepreneurs see AI as complex or poorly aligned with their business needs, often due to limited understanding of what problems AI can realistically solve and how to implement it step by step. A lack of managerial understanding and technical talent can delay AI implementation for months or even years, especially in smaller firms.
Women are about 25% less likely than men to adopt AI, a gap driven largely by confidence and familiarity rather than a lack of interest or ability. Young people report widespread use of AI tools but say training on ethics and responsible use is insufficient.
- Data access shapes outcomes
AI systems depend on high-quality data. In many developing countries, relevant local and sector-specific data are scarce, fragmented or costly. This limits entrepreneurs’ ability to tailor AI tools to local markets and business realities. Open data initiatives and data-sharing frameworks can help close this gap, while maintaining privacy and trust.
- Regulation can enable or constrain innovation
Clear, proportionate and predictable regulation is essential. Rules designed for large firms can overwhelm MSMEs and start-ups. The report highlights the value of risk-based approaches, sector-specific rules aligned with national development priorities and regulatory sandboxes that allow firms to test AI solutions safely while regulators learn alongside innovators.
- Finance remains a binding constraint
Adopting AI requires upfront investment in skills, software and infrastructure. Many MSMEs struggle to access credit or equity financing. Blended finance, public guarantees, targeted subsidies and fintech solutions can lower risks and expand access to capital, helping AI adoption reach smaller firms. They can also support phased approaches to adoption, from off-the-shelf tools to partnerships and, eventually, in-house capabilities.
Policy recommendations
Building on these findings, the report provides policy recommendations to help entrepreneurs adopt AI – focusing on:
- Skills development
- Access to data and finance
- Innovation-friendly regulation
- Stronger entrepreneurial ecosystems