The Human Feedback Foundation, with support from RBC, has launched the Agentic AI Design Exchange. It is described as the first national program to help Canadian nonprofits design, build and use their own dedicated AI agents. Select designs will go into an open library that other nonprofits can copy and build on.
What the program does
An AI agent is a tool that can take on tasks, not just answer questions. Under the program, teams from nonprofits will find a real need in their community and build an agent to address it. They will put fairness, accountability and human oversight into the design from the start.
The method comes from Professor Mihnea Moldoveanu of the Rotman School of Management at the University of Toronto. The program builds on an earlier effort, Responsible AI for Social Impact, known as RAISE.
Who is behind it
The Human Feedback Foundation is a federally incorporated Canadian nonprofit. The Exchange is funded by a community investment from RBC through its Stronger Nonprofits Program. Other partners include the Rotman Business Design Initiative, Imagine Canada and the Canadian Centre for Nonprofit Digital Resilience.
Why the timing matters
Nonprofits are behind on AI. A 2026 benchmark from Imagine Canada and the Canadian Centre for Nonprofit Digital Resilience found that only 20 per cent use AI for more than three activities. Most have no formal AI policy.
Business is moving faster. A KPMG survey cited in the announcement says nearly six in ten large organizations are deploying or scaling AI agents.
Who it affects
- Nonprofit staff and volunteers who want help with AI but lack a plan.
- Small charities that could reuse shared designs instead of starting from scratch.
- People who rely on nonprofits for services.
What is not known yet
The first group of nonprofits is scheduled to begin in the fall of 2026. The announcement does not say how many teams will join, how to apply, or when the first designs will appear in the open library. More information is promised at humanfeedback.io.
The takeawaySmall nonprofits may soon be able to borrow free, open AI designs that already include safety and human oversight.