UN report warns AI could deepen inequality between rich and poor without swift action

Artificial intelligence could exacerbate global inequality unless urgent measures are taken to ensure its benefits are shared more widely, the United Nations Development Programme has warned in a new report. The analysis argues that while AI holds immense promise, most gains are likely to accrue to wealthier nations unless governments invest in closing gaps in basic infrastructure, digital access and skills.

The report, cited by AP, released on Tuesday, likens the moment to the Great Divergence of the industrial revolution, when rapid modernisation in Western economies left many others behind. It notes that debates around AI largely focus on productivity, competitiveness and growth, whereas the more fundamental question is how the technology will affect human lives—particularly those already marginalised in data-driven systems.

The UNDP highlights that communities lacking reliable electricity, internet connectivity and education—along with older people, those displaced by conflict, and populations affected by climate disasters—risk being excluded from the AI economy. Many such groups are effectively invisible in datasets, meaning the systems built on that data may fail to account for them.

AI’s potential to lift productivity and accelerate development remains significant, the report says, citing applications such as improved farming advice, faster medical diagnoses, enhanced weather forecasting and more accurate damage assessments. These advances could transform rural areas and disaster-prone regions. AI technologies that analyse poverty, health and disaster risks can also support quicker and more transparent decision-making, turning data into public value.

However, even advanced economies face challenges. The growing demands of data centres for electricity and water have prompted concerns that increased power generation could slow progress on cutting carbon emissions and aggravate health hazards. Ethical, privacy and cybersecurity risks also persist, with researchers observing hackers using AI to automate aspects of attacks, alongside the rising threat of deepfakes enabling misinformation and criminal activity.

The report notes that countries such as China, Japan, South Korea and Singapore are well positioned to harness AI, while others—including Afghanistan, the Maldives and Myanmar—lack the power infrastructure, skills and resources needed to participate. Regional inequalities within countries could further widen the divide, even in technologically advanced economies.

About a quarter of the Asia-Pacific region remains offline, raising the prospect that millions could be shut out of digital payments, digital IDs and education systems essential for engagement in an AI-driven global economy. Without intervention, the report warns, many may find themselves stranded on the wrong side of a rapidly evolving technological landscape.

It emphasises risks linked to surveillance, opaque systems that entrench biases, and widespread misinformation, underscoring the need for transparency and strong regulatory safeguards. AI is fast becoming essential infrastructure on par with power and roads, the report states, but with both greater benefits and sharper risks.

Governments are urged to invest heavily in digital infrastructure, education and training, fair competition and social protection. The goal, the UNDP stresses, is to democratise access to AI so that every country and community has the opportunity to benefit while shielding the most vulnerable from disruption.