AI for Marginalized Communities!

 AI is a buzzword worldwide—but who truly benefits from it? And how do marginalized people perceive it? Answering these questions isn’t easy.

Before seeking solutions, we must first understand the current state of AI adoption and its impact on marginalized communities, particularly people with disabilities.

Take Mahafuza, a 36-year-old woman from Rajshahi District, Bangladesh. She has a disability and participates in a Focus Group Discussion organized by UNDP Bangladesh in January 2025 to assess the country's AI readiness. When asked to describe AI in one word and express it through painting, she created a striking image.

Can you guess what it represented? After she finished, I asked her to explain. She said, “AI is something dark. Although it’s technology, it doesn’t seem good to me—maybe even bad.”

What do you think? Could AI widen the digital divide for marginalized communities? If so, what solutions are needed for underserved populations?

The real challenge isn’t just finding a solution — it’s identifying the actual problem. How do we ensure AI solutions are truly demand-driven and address the real needs of marginalized groups?

The key issue is not about "I" or "WE," but THEY. If communities themselves cannot define their challenges—if they struggle to articulate the barriers keeping them in the digital divide—then real solutions remain out of reach.

Too often, solutions are proposed before fully understanding the problem. This is a critical mistake that prevents sustainable progress. The real challenge isn’t just a lack of resources but ineffective policies and flawed approaches to inclusion.

Let’s continue the discussion.

Key Insight: A citizen demand-driven approach is essential to identifying pain points and fostering sustainable solutions for marginalized communities.