AIFOD Review Examines Global AI Divides and Developing-Nation Priorities

Developing nations must prioritise foundational infrastructure and basic access over frontier technology to avoid entrenched technological inequality, according to experts speaking at Session 33 of the Weekly AI Review organised by the AI for Developing Countries Forum (AIFOD), which publishes this site. The panel explored a four-tier framework originating from an estimate by Andrej Karpathy, which contrasts roughly six billion individuals who have barely interacted with artificial intelligence against the few million delegating complete workflows to automated agents.

Infrastructure Before Frontier Models

Panellists argued that closing the divide between simply conversing with chatbots and deploying operational agents does not require leading-edge models. Dr Alan Ross, head of the International Association of AI-agent Developers, noted that developing economies can achieve significant gains using existing systems from China or elsewhere, provided sovereign data concerns and data centres are addressed. Instead of pursuing top-tier development labs, contributors emphasised building domestic capacity while keeping human supervision in place.

Dr John Hemery of the Centre for Political and Diplomatic Studies urged authorities to facilitate fundamental prerequisites rather than race ahead with premature regulation. He stressed that reliable electricity, municipal internet access and telecommunications should precede advanced deployments, citing the Western Cape provincial administration's Digital Transformation Plan 2.0 in South Africa as an example of integrated connectivity.

Economic Disparities and Governance

Industrial experts highlighted substantial funding gaps affecting adoption. Capgemini vice president Sudeep Rudra pointed out that leading global organisations outspend median enterprises by up to thirty times, advocating public computing subsidies similar to India's digital public infrastructure. Meanwhile, Whitelibra founder Nicolò Boggian warned that historical patterns risk repeating, where workers in poorer territories contribute to model training while commercial revenues concentrate in Western markets. The panel concluded that overcoming these divides requires urgent policy planning, structural workplace transformation, and educational reform across emerging regions.