Recently, the World Bank Group (WBG) released its flagship report on Artificial Intelligence (AI) 2026, titled "The Promise of Artificial Intelligence." It is the institution's first detailed assessment of how people, businesses and governments in poor countries are using AI, what is holding them back, and where the largest gains are likely. The report offers a simple framework for decoding AI and striking a sensible blend of measures to adopt, adapt and advance AI-based technologies, especially for developing countries. This speaks to Africa directly.

The report frames AI as a historic opportunity for developing countries, offering a roadmap for harnessing "small AI" tools in health, agriculture and education, while warning of risks such as dependency and inequality. On the brighter side, it argues that developing economies have more to gain and less to fear from AI than richer ones. It notes that less than a tenth of existing jobs in low- and middle-income countries, where most African countries lie, are at risk of automation by generative AI, compared with more than a third in high-income economies. On this point, it goes on to emphasise that  AI, for example, could enable judiciaries to radically reduce case backlogs, help teachers prepare better lessons, assist nurses in interpreting medical scans, or guide a farmer on when to plant and which pests to avoid.

It cites Kenya's Judiciary as a case study on its deployment of AI tools to assign more than 10,000 court cases each year to over 1,500 mediators, helping clear a large backlog. On the political front, it points to Ukraine, where a machine-learning tool built by Transparency International Ukraine detects suspicious patterns in government procurement and channels citizens' complaints to law enforcement, a model African anti-corruption agency would do well to study.

The report sets out pathways for developing countries such as: adopting existing AI tools off the shelf, adapting them to local needs, and advancing frontier technology by building homegrown models and infrastructure. However, it cautions developing nations that the path of advancing new frontier models is by far the hardest and most expensive. It demands chips, large data centres, vast training data and world-class researchers-resources only a handful of companies in a few developed countries can currently afford, spending billions to train leading models. For most African states, the pragmatic starting point lies firmly in adopting and adapting models developed by developed nations.

This is where the legal and governance dimension becomes decisive. The report is candid that AI's benefits are not automatic: gaps in infrastructure, data, skills and institutional capacity could widen existing inequalities, while weak governance may amplify risks ranging from algorithmic bias to market concentration. Its counsel to governments acting as regulators is measured rather than alarmist: start with voluntary industry standards as a foundation for trustworthy AI, then lean on existing consumer-protection, competition and data laws to address concrete harms such as unsafe systems, discrimination and surveillance, rather than rushing to draft entirely new AI statutes. For Uganda and its neighbours, many of whose data-protection and cyber laws are still young, this is a call to strengthen enforcement of what already exists before chasing novel legislation.

Governments also feature as users of AI, not merely referees of it. The report urges public agencies to deploy AI in service delivery, tax administration, social protection, disaster response, health and education, but insists this requires making public administrative data "AI-ready" and building frameworks for testing, procuring and evaluating AI solutions before scaling them. Public procurement, it argues, can be wielded strategically to crowd in local AI developers where imported tools prove ineffective or harmful in high-stakes settings such as healthcare or the courts. This is a direct opening for African software developers and legal-tech innovators, provided procurement rules are constantly updated to let them compete.

Crucially, the report frames sound governance not as red tape but as the price of public trust. It calls on states to protect citizens from discrimination, surveillance and misinformation, to invest in the electricity, connectivity and digital skills that make adoption possible in the first place, and to participate actively in international AI governance rather than leaving the rules to be written solely by the handful of countries and companies that dominate the technology. Ideally, the point should resonate loudly across the continent: Africa's digital sovereignty will be shaped as much by its seat at global standard-setting tables as by any domestic law it passes.

Ultimately, the WBG casts its report less as a prediction than as a development strategy for living with AI. Its message for Uganda, Kenya and the wider region is unambiguous: treat AI as one tool within a broader project of institutional and economic development, not a shortcut around it. Closing the gaps in power, connectivity, skills and governance capacity will determine whether African countries merely consume this technology or begin, in whatever modest ways, to shape it.