World Bank Pushes 'Small AI' Models for Developing Nations
The World Bank says lightweight AI designed for low-connectivity mobile devices could compress centuries of economic development into decades.
The World Bank is calling on developing countries to build and deploy 'small AI' models — streamlined artificial intelligence tools capable of running on a smartphone without reliable internet access — as a means of accelerating economic growth at a pace that traditional development pathways cannot match.
Paschal Donohoe, the institution's chief knowledge officer, argued that while major Silicon Valley companies remain focused on increasingly powerful and data-hungry consumer AI products, a parallel opportunity exists for emerging economies to harness leaner, locally adapted versions of the technology. The distinction matters: general-purpose AI systems built for high-bandwidth environments in wealthy nations may offer little practical utility to communities with limited digital infrastructure.
Read more Oxford Street Goes Car-Free: Will Pedestrianisation Revive London's Busiest Strip? →
Donohoe framed the potential impact in sweeping terms, suggesting that appropriately tailored AI could compress economic development timelines from centuries down to decades — a claim that underscores how urgently the World Bank views the window of opportunity. The key condition, however, is that the technology be adapted to local languages, contexts, and connectivity constraints rather than imported wholesale from abroad.
The institution's push reflects a broader recognition among development economists that the AI revolution risks deepening existing global inequalities if low- and middle-income countries are left dependent on tools built by and for wealthier markets. Encouraging domestic or regional 'small AI' development would give those nations greater control over how the technology shapes their public services, agriculture, healthcare, and financial systems.
Continue reading at Business | The Guardian.
