One participant in our community learning programme ran a simple test. She typed the same disclosure of domestic violence into four different AI models and compared the answers. Only one recognised that the person writing needed both a safe house and specialised GBV support. In her words, “model selection stopped feeling like a technical decision to me and became a safety decision.”
Moments like this are happening across women’s rights organisations in Africa. They rarely make headlines, yet they are where feminist AI becomes real: in the choices organisations make about which tools to trust, which data to collect, and when to say no. At the A+ Alliance, we want these stories to be heard. That is why we are launching GAIC Coffee, a new series on the AI & Equality Community Circle space that brings them into the open.
GAIC Coffee is an informal, welcoming space for the women’s rights organisations that took part in the GAIC Community Learning Programme, which the A+ Alliance ran in two cohorts in 2026. During the Community Learning, participants explored feminist AI and decoloniality, human rights approaches to technology, and how bias enters AI systems, working through the full AI lifecycle from framing a problem and collecting data to selecting and auditing models. GAIC Coffee invites them back to share what happened next: how these principles have shaped their projects, how they have brought a human rights approach into their organisations and communities, and how they are noticing and addressing bias in the tools and systems around them.
The Gender & AI Innovation Collective (GAIC) Community Learning Programme drew 552 applications across its two cohorts and welcomed 234 participants from women’s rights and civil society organisations. Many arrived seeing AI as a finished product to adopt, a “black box” built elsewhere by others. The programme was designed to open that box. Each module paired concepts with hands-on exercises, and take-home assignments asked participants to work through the needs of their own communities, so the learning was anchored in their own contexts from the start.
Three convictions that travelled home
When participants described how their thinking changed, three convictions came up again and again.
Responsible AI begins by asking whether AI is needed at all. Participants learned to treat “no” as a valid and valuable answer, and to weigh a tool’s promise against the risks it carries for the people it is meant to serve.
Data collection is an exercise of power. Who is counted, who is left out, and who decides are the places where bias and exclusion get built into a system. Participants now approach data with that awareness.
Communities are experts in their own lives. Tools designed for women and girls should be shaped and owned by them, which is why many organisations have made co-design a firm requirement.
These convictions are already changing practice. Organisations have paused or redesigned tools that could put survivors and patients at risk, built privacy safeguards and community oversight into their systems, and left the programme with working concepts, revised designs, and pilot plans. Alongside new skills came a growing sense of authority. As one participant from Mali put it, the programme gave her organisation “the legitimacy and arguments to defend this vision.”
Gender & AI Innovation Collective today:
The Gender & AI Innovation Collective puts women’s rights organisations in the design seat of AI. Co-led by Women at the Table and Code for Africa, it supports organisations to understand how AI systems work, decide whether AI truly serves their communities, and build tools they own inside the AI Innovation Sandbox.
Seven organisations across six countries are now building. These teams will present their work in public showcases on Circle in January and February 2027. GAIC Coffee runs alongside this work, widening the circle to include every organisation that carried the learning home. The first session will be on November 20th.
Join us to hear directly from organisations across Africa, ask your own questions, and see what feminist AI looks like when communities lead it. Pour yourself a coffee, we’ll save you a seat.
Photo: Elise Racine / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/