African and European researchers are being urged to shape AI for science together, with locally representative data, shared research infrastructure and African leadership at the centre. The proposals are an attempt to ensure the technology serves research needs on both continents, rather than simply extending the reach of systems built elsewhere.
Watch Desk analysis
What happened
At an African Union-European Union meeting on AI in science on 26 October, researchers and policymakers discussed cooperation on applications from public health and food security to climate and biodiversity. The gathering moved online at short notice after a planned ministerial meeting was postponed until 2027; political consideration of the group’s recommendations is still to come.
The meeting’s proposals include connecting research repositories, sharing computing facilities and evaluating AI tools in the settings where they will be used. Participants also called for collaboration to build on existing Africa-Europe research networks, with stronger links between universities, research organisations and businesses to help turn findings into products and services. Read the University World News report.
Why it matters
AI for science depends on more than access to a model. Researchers need computing resources, useful data and the ability to test whether a system works for the populations and environments it is meant to serve. At the meeting, South Africa’s UNESCO delegate Phil Mjwara argued that gaps in genomic data can limit AI-assisted drug discovery for genetically diverse populations. As he put it, a model cannot learn biology missing from its training data.
The discussion also recognised that infrastructure alone will not close the gap. Skills, research careers, funding and fair access to data all affect who can contribute to scientific work and benefit from it.
Our read
This is a substantive agenda, not a signed infrastructure deal: the recommendations still await political consideration. Its strongest point is that inclusion is not an optional flourish if the science itself depends on whose data are represented. The next test is whether the partnership turns calls for shared resources and African leadership into funded, practical commitments.
What to watch
- Whether the postponed ministerial meeting adopts concrete recommendations.
- How joint computing and data access are funded and governed.
- Whether research projects set measurable goals for locally representative data and African leadership.
Discussion spark: Should Africa-Europe AI research prioritise shared infrastructure, or put locally governed data and research leadership first?
Sources and evidence
- Africa and Europe seek their own path in AI science (5 November 2026, 00:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.