#ShiftThePower and AI: Who wants to talk?
30 Aug 2026
Matt Haikin, Possible Now Stitching Fellow
As one of the new Possible Now Stitching Fellows, I will be joining the other fellows and hundreds of others from the global #ShiftThePower community in Kathmandu for the Possible Now Summit.
It is vital that AI forms a part of that conversation, given the impact it is already having in the world – for good and for ill.
I plan to arrive with something more relevant than the typical “there are opportunities but also risks of harm” or “it’s complicated”, by hosting a series of conversations between now and November with another Fellow, Facundo Ibarlucía of Red Comunidades Rurales in Argentina.
The broad question we are posing in these conversations is: What does a #ShiftThePower approach to AI look like?
Please join one of those conversations by completing this short form and we will contact you with details of the various conversations planned between now and the end of November.
You absolutely do not need to be an AI expert to join!
Why this question, and why now?
There is a wealth of advice on using AI tools effectively, and almost as much warning against AI’s potentially devastating impacts on society, but comparatively little that is nuanced about harnessing AI for good while minimizing or mitigating (though rarely removing entirely) its negative impacts.
Add an explicit power-dynamics lens and a focus on locally-led development, and the advice dwindles further.
Most social impact organizations and funders are already experimenting with AI. If questions of power, ownership and agency are going to shape the role of AI in global development, these conversations need to start now.
For me, the bigger #ShiftThePower question above breaks down into two very different parts:
- How can the #ShiftThePower community take advantage of the opportunities of AI to further its goals? In other words: how do we maximize its usefulness in furthering the mission, while minimizing associated harms and avoiding exacerbating the power inequalities already inherent to AI?
- Is there a role for the #ShiftThePower community in directly addressing AI’s own power inequalities? For example, by influencing global debates, challenging some of the assumptions and incentives shaping AI, or supporting the development and use of alternatives to those provided by “Big Tech.”
Those questions open up plenty of others: What happens when mainstream AI performs badly in particular languages, cultures or contexts? What happens to philanthropy and trust when AI is both writing and scoring grant proposals? Can smaller, locally hosted or locally developed AI genuinely shift power — and, if so, which bits of power? When is choosing not to use AI the better option? What are others already doing that we should know about?
That is very much in the spirit of the Possible Now Summit, which is about asking big systemic questions rather than assuming we already know the answers. The Stitching Fellowships are about bringing together perspectives, ideas and experiments that are often happening in different places and different conversations. AI is definitely one of these big questions and our aim is to stitch together different perspectives on it before December, and bring something richer and more grounded into the conversation in Kathmandu.
How will the conversations work?
We want to listen to the experiences, needs, concerns, promising practices and failures of diverse practitioners — and particularly people living with the consequences of the choices organizations are making around AI.
The aim is not to produce a neat consensus position on whether AI is good or bad.
Instead, we want to start joining the dots:
- Where are people seeing genuine value?
- Where is AI reinforcing existing inequalities or creating new ones?
- What choices are organizations making about data, ownership, infrastructure and control?
- What experiments are worth learning from?
- Where do people strongly disagree?
- And what questions are we still not asking?
We will stitch what we hear together and take it into an interactive AI Lab at the Summit in Kathmandu — not a staid summary but lived experience, questions, tensions and possible approaches for the wider #ShiftThePower community to engage with in person. Watch this space for more info on the Lab as it develops (and contact Facundo and I if you would like to help us shape it!).
Who should join?
We would particularly like to hear from:
- People working on locally-led development, community philanthropy or shifting power;
- People from regions and languages badly served by mainstream AI;
- Technologists building or deploying local and / or open AI models or tools;
- People working on digital rights, data governance, technology sovereignty or related areas;
- People already experimenting with AI who have something useful — or disastrous — to share;
- People deliberately refusing to engage with AI, and able to explain why.
Experiments, failures and strong opinions are particularly welcome.
You do not need to bring expertise in AI. Healthy curiousity, lived experience and asking hard questions can be just as useful.
If you understand AI but have never had to think hard about power, join in. If you have spent your career on power shifting and localization but feel the AI conversation is moving too fast to get into, join in. And if, like me, you are somewhere in the awkward overlap, definitely come along.
The series will run from now until the end of November. Please complete this short form to register your interest, and we will invite you to one of the upcoming conversations.
We hope to arrive in Kathmandu with a richer set of experiences, disagreements, experiments and questions than any of us could bring on our own.
By: Matt Haikin, Possible Now Stitching Fellow