The Global AI Summit on Africa #GAISAfrica2025 | Kigali, 3 April 2025
KIGALI — Rwanda can use government spending to help build its local artificial intelligence industry, the World Bank says in its latest global study on AI adoption.
The Bank points to Rwanda’s Public Procurement for Innovation (PPI) framework as an example other developing countries could learn from. The approach allows government agencies to describe a public problem without prescribing the technology needed to solve it.
Companies can then compete by proposing and developing solutions in stages, while retaining intellectual property rights. This could give start-ups and local innovators a better chance to compete with large international companies.
Governments are already among the biggest buyers of goods and services in developing countries. The World Bank says this gives them an opportunity to create demand for local technology and help new companies grow.
Adapt AI rather than build it from scratch
Rwanda also stands out in its broader approach to AI.
It is the only one of the world’s 25 low-income countries to have published a national AI strategy, according to the World Bank.
More than 80 countries globally had published national AI strategies by mid-2026, putting Rwanda ahead of its low-income peers in setting a formal direction for AI development and use.
In ‘World Development Report 2026: The Promise of Artificial Intelligence’, released this week, the World Bank says developing countries should not try to build expensive frontier AI systems from scratch.
Instead, they should focus on using existing AI, adapting it to local needs and building their own capabilities where necessary.
Public procurement can support this process. A health agency, for example, could allow a local company to test an AI tool for disease diagnosis, while a tax agency could test AI for document processing before deciding whether to buy the technology at scale.
The Bank also says procurement rules can require systems to be interoperable, auditable and portable, helping governments avoid becoming locked into one technology provider.
This can turn government into an anchor customer, helping local companies move from ideas and prototypes to commercially viable products. For countries with limited resources, the Bank says this could be one of the most practical ways to build local technology capability.
Rwanda puts innovation procurement into law
Rwanda has moved to formalise the approach.
In February 2026, the Ministry of Finance and Economic Planning issued a ministerial order establishing PPI within the country’s procurement system. The rules allow public institutions to buy innovative ideas and prototypes rather than only finished, off-the-shelf products.
The framework builds on pilots run by the Rwanda Information Society Authority since 2024, including an AI-powered procurement platform and a digital research portfolio system developed by a local start-up.
Under the new rules, agencies can use design contests, where they present a public problem and invite companies to propose solutions. They can also use pre-commercial procurement, which funds early research and development before the strongest solutions move toward full contracts.
The rules also remove some barriers that traditionally favor larger companies, including heavy bid guarantees and strict previous-experience requirements.
Implementation will be the real test
The World Bank cautions that procurement can also fail if poorly designed. It can strengthen established companies and lock governments into one supplier instead of opening opportunities for new firms.
For Rwanda’s model to work, procurement officers will need the skills to assess innovative technologies, evaluation committees will need to remain independent, and successful pilots will have to prove they can work at scale.
Rwanda has become one of the first African countries to formally put innovation-focused public procurement into law. But the bigger test is whether the policy creates a competitive local AI industry, or simply produces government-funded projects that never become commercially viable.
Ultimately, the success of PPI will be measured not by the strength of the policy on paper, but by how many useful local companies and products emerge from it.
