Building and Scaling AI for Social Impact
Event Highlights
Kalpa Impact and The Rockefeller Foundation convened a workshop, at The University of Chicago Center in Delhi, during the India AI Impact Summit 2026.
The invite-based workshop convened over 50 participants across the funding and tech ecosystem, think tanks, grassroot organisations, and academia to explore the constraints and opportunities in building and scaling AI for social impact.
The session saw opening remarks by Deepali Khanna – Senior Vice President and Head of Asia, The Rockefeller Foundation, and Andrew Sweet – Vice President, Innovation, The Rockefeller Foundation.
Andrew Sweet opened the workshop with a reflection: “In August 1955, The Rockefeller Foundation received a bold and prescient proposal from a young mathematician named John McCarthy at Dartmouth College. He wanted support for a two-month summer study “to find how to make machines use language, form abstractions and concepts, solve problems now reserved for humans, and improve themselves.” In that proposal, which The Rockefeller Foundation supported with a USD 7,500 grant, McCarthy coined a new term: artificial intelligence. Seventy years later, AI is here, powerful, pervasive, and reshaping nearly every corner of society.”
“This group covers the entire ecosystem, but rarely gets an opportunity to simply sit down and talk with each other. Tonight, we wanted to enable that” – Deepali Khanna set a prescient note for the workshop.
The workshop, facilitated by Sushant Kumar – Founder and CEO, Kalpa Impact, had a unique two part design, with the first half being curiosity mapping, smaller more intimate discussions with a curated table assignment. Followed by a marketplace format where participants could discuss with broader groups interested in similar thematic areas.
Participants surfaced their unique challenges and forged new partnerships in the following thematic areas:
- Data availability and quality
- Talent and capacity
- Trust and accountability, governance
- Compute access and affordability
- Inclusion
- Efficiency to scale (financial viability)
- Green compute and other open areas
What Comes Next
The event was designed to provide a space for curiosity and exploration to thrive. Its design was not meant to offer silver bullets or quick fixes to deeply embedded challenges.
AI for social good requires continued exploration between those who build, those who enable, and those who sustain. It is rare that all of these parties are present at the same table. By bringing together a curated set of individuals and crafting a space for cross-sectoral dialogue, honest reflection, and hopeful provocation, we attempted to enable the ecosystem.
Mapping the 45 curiosity cards that participants filled during the event surfaced one thread: the evening led to connections that will go beyond the workshop in the form of pilots, working groups, or even just soft connections. It also made us realise there is appetite in the ecosystem for more such spaces and conversations, and we plan to enable the same in the future.