View noticesAI market set to hit $4,8 trillion by 2033, but benefits remain uneven
The global artificial intelligence market is projected to reach $4,8 trillion by 2033 – roughly the size of Germany's economy – marking a 25-fold increase over just ten years, according to the UN Conference on Trade and Development (Unctad).
The agency said this rapid expansion is already shaping everyday life in developing regions. AI-powered mobile apps are helping rural health workers interpret symptoms in areas where doctors are scarce, adaptive learning platforms are personalising lessons for students in under-resourced schools, and in climate-vulnerable regions, AI models are forecasting floods, droughts and crop failures days or weeks in advance, giving communities time to prepare.
However, Unctad cautioned that these gains are not automatic, and depend on decisions about who builds the technology, whose data trains it, and who has a say in how it is governed.
A widening divide
The benefits of AI remain far from evenly distributed, the agency said. A small group of economies and around 100 firms control the bulk of global AI research, patents and computing power. Fewer than a third of developing countries currently have a national AI strategy, and among least developed countries that figure drops to just 12%.
Roughly 65% of people in least developed countries remain offline, putting AI out of reach before they even begin. At the same time, Unctad estimates that up to 40% of jobs worldwide could be affected by AI-driven automation, a shift that threatens to erode the low-cost labour advantage many developing economies rely on for growth and poverty reduction.
The agency also flagged a governance gap: 118 countries, mostly in the global South, are absent from major AI governance forums, meaning international norms on safety, transparency and accountability are being written without their input.
Unctad said these were not merely technical shortfalls, but human development gaps affecting livelihoods, public services and countries' ability to meet their citizens' needs.
Where AI could have the greatest impact
Unctad said AI could deliver the greatest human development impact in areas where needs are most acute – agriculture, primary healthcare, disaster management and public administration.
Success in these areas depends on three interdependent factors: infrastructure such as electricity, broadband and computing power; locally relevant data; and workforce skills.
For most developing countries, the agency said, the realistic entry point is adapting existing or open-source models to local conditions, rather than attempting to build proprietary systems from scratch.
Aligning these efforts with national development plans, and treating data governance as central to AI readiness, is what Unctad said would separate countries able to capture AI's benefits from those where the technology remains confined to urban centres.
Policy recommendations
Unctad argued that an inclusive approach to AI is not a slower path, but one that ensures the technology leaves no one behind, adding that the window for developing countries to act is open but narrowing.
The agency called for global cooperation to steer AI development towards shared goals and values, setting out four key priorities:
Industry commitment – a public disclosure mechanism for AI, similar to existing environmental, social and governance (ESG) frameworks, to improve accountability and translate global commitments into concrete outcomes.
Shared infrastructure – a global shared facility to provide equitable access to AI infrastructure, lowering entry barriers for countries unable to build such infrastructure on their own.
Open innovation – expanding access to open data and open-source models to support inclusive AI innovation, including better coordination of fragmented open-source resources.
Capacity building – sharing AI knowledge and resources, particularly through South–South cooperation, to strengthen developing countries' capacity to benefit from AI and address shared challenges.