UNCTAD Calls for Urgent AI Ecosystems in the Developing World
Geneva: As artificial intelligence (AI) reshapes economies worldwide, developing countries face a stark choice: rapidly develop robust ecosystems to help their small businesses and startups embrace the technology, or watch existing global disparities grow even larger. This urgent warning forms the centrepiece of a detailed new report from the United Nations Conference on Trade and Development.
The report, “Artificial Intelligence Unleashed: Transforming the Entrepreneurial Scene in Developing Countries,” examines the specific opportunities artificial intelligence presents for micro, small, and medium-sized enterprises (MSMEs)—as well as for innovative startups in these nations. At the same time, it lays out the equally specific barriers that frequently prevent them from seizing those opportunities.
Prepared with contributions from a wide array of sources, including in-depth reviews of existing research, direct consultations with representatives from international organisations, government agencies, private companies, nonprofit groups, investors, academic institutions, AI specialists, and entrepreneurs themselves, the report also incorporates findings from surveys conducted at more than 40 centres within UNCTAD’s global Empretec entrepreneurship training network. The conclusion is straightforward yet profound: smaller enterprises dramatically increase their chances of successful AI adoption when embedded in a supportive ecosystem. Key elements of this ecosystem include regulatory frameworks that encourage responsible experimentation without excessive burdens, accessible platforms and practical tools, dedicated programmes to build both long-term strategic thinking and immediate technical capabilities, and financing arrangements that reduce the financial risks and upfront costs associated with trying new technologies.
The potential benefits AI can bring to these businesses are substantial and multifaceted. It enables automation of repetitive tasks, optimisation of procurement and inventory management, significant overall cost savings, improved operational efficiency, more targeted marketing efforts, highly personalised customer experiences, and the cultivation of an internal culture focused on continuous innovation and environmental sustainability. Importantly, these advantages do not appear automatically. They emerge only when a business’s internal capacities align well with favourable external conditions—an alignment that varies greatly depending on location and company scale.
Firms in advanced economies benefit from superior digital infrastructure, thriving innovation communities, and abundant skilled talent, allowing them to adopt AI more readily. In developing contexts, however, enterprises often confront deep-rooted deficiencies. Entrepreneurs frequently lack a clear understanding of how AI can address their particular operational challenges. Digital infrastructure remains inadequate. Access to high-quality data—essential for training or effectively applying AI models—is limited. Larger firms enjoy advantages in vast datasets and powerful computing resources that smaller ones simply cannot match, creating a cycle where the technologies most reliant on scale end up benefiting those already in strong positions.
A particularly hopeful aspect explored in the report is AI’s capacity to broaden entrepreneurial participation, especially for groups historically underrepresented, such as women and youth. Studies cited show women entrepreneurs engaging with AI roughly 25 per cent less often than men, largely because of differences in confidence and familiarity. Many young people, by contrast, already use AI tools regularly but express a strong desire for more comprehensive training on ethical issues. Well-designed initiatives targeting these groups could significantly expand involvement. Such efforts, however, must overcome pervasive connectivity problems: while internet access approaches universality in high-income countries, it reaches only about 54 per cent in lower-middle-income nations and drops to 27 per cent in low-income ones, with even lower figures in least developed countries. The primary causes—lack of physical infrastructure, unaffordability, and insufficient basic digital literacy—leave millions disconnected and unable to participate.
Governance plays a crucial role, requiring active collaboration among diverse stakeholders to develop AI that is safe, secure, trustworthy, and aligned with varying national values and global ethical norms, emphasising human rights, transparency, and accountability. The report delves into challenging questions, such as how to clearly define responsibility when AI systems produce unintended outcomes, the importance of narratives that spark business creativity while minimising risks, including algorithmic bias, discrimination, intellectual property infringements, and opaque decision-making processes, and the perpetual need to harmonise regulatory oversight with the pace of innovation.
Promising approaches include designated regulatory sandboxes for controlled, low-risk experimentation, promotion of open-source tools customised to local languages and contexts, innovative financing models such as grants, risk-sharing funds, and incentives to attract private investment, and expanded educational offerings that emphasise practical applications—for instance, using conversational AI for customer support or predictive analytics for supply planning.
Concrete examples bring these concepts to life. Certain logistics companies in Malaysia and Saudi Arabia have successfully applied AI to streamline cross-border fulfilment and baggage handling, demonstrating measurable efficiency improvements. These cases, however, remain exceptions rather than the rule, typically occurring only where multiple supportive factors coincidentally align.
A looming financing shortfall for MSMEs, projected to reach $2.6 trillion globally by 2030, poses another major hurdle. The report proposes remedies, including subsidies, de-risking instruments, targeted educational funding, and mechanisms to encourage greater private sector involvement.
The publication concludes with an adaptable policy framework drawn from UNCTAD’s established Entrepreneurship Policy Framework. It organizes recommendations into five interconnected pillars: adapting regulations to provide clear standards, defined accountability, transparency requirements, and collaborative processes including sandboxes and training for officials; enhancing education and skills through public-private partnerships that cover technical, ethical, and practical dimensions; facilitating technology diffusion by having governments demonstrate AI use in public services, promoting customizable open-source options, creating resource hubs, and fostering linkages between large and small entities; improving access to finance via dedicated public funds, tax incentives, subsidies for tools and training, and risk-mitigation strategies; and promoting awareness and networking through campaigns highlighting benefits and ethics, showcases of successful implementations, and peer platforms to reduce apprehension and encourage hands-on exploration.
Overall, the publication asserts that alignment of core components—reliable linkages, learning opportunities, funding channels, and appropriate direction—positions AI to drive context-specific creations tackling immediate community concerns, bolstering market presence, and advancing wider aims for enduring progress. When good digital infrastructure gives everyone smooth internet access, training programs teach entrepreneurs the technical skills and ethical knowledge they need, financing options make it easier to get started with targeted help, and clear but flexible rules guide the way, small businesses feel brave enough to try out AI tools and improve them over time. These tools can solve real local problems, like helping farmers in remote areas grow more crops by predicting the best planting times or making healthcare better in places with few doctors. This teamwork of support does more than help single companies compete at home or abroad. It builds stronger national economies by creating new jobs, using resources wisely, and growing in ways that include everyone—matching big goals for long-term, planet-friendly development. These connected systems also spark original ideas from within the country, not just copying tech from elsewhere. Developing nations can then create special strengths based on their own cultures and daily needs. But if countries do not act quickly and together to build these supports, the fast growth of AI will likely make old gaps bigger. What could be a chance for everyone risks becoming another field where only the powerful win.
– global bihari bureau
