Environmental justice / citizen scienceenvironmental justice and air quality advocacyParticipation documented

People-centered AI for Amazon coastal resilience

Curuçá, Brazil

Rare describes the pilot as combining local knowledge, low-cost technology, and AI in a people-centered, human-in-the-loop approach for community resilience.

participatory designcommunity governanceproblem formulationdata collectionmodel developmentdeployment

Project documentation

Overview

Region: Latin America
Lead organization: Rare / Microsoft AI for Good Lab
Organization type: nonprofit / NGO
Technology group: AI-assisted natural resource governance
Activity status: pilot
Start year: 2024
Last updated: Mar 25, 2026

Participation documentation

Participation mode: community-led participatory design and governance
Participants: coastal communities; local fishers; scientists; NGOs
Methods: community observation; human-in-the-loop analysis; local knowledge integration
AI lifecycle stages: problem formulation; data collection; model development; deployment
Evidence summary: Rare describes the pilot as combining local knowledge, low-cost technology, and AI in a people-centered, human-in-the-loop approach for community resilience.
Atlas assessment: Core · High confidence · evidence grade A
Uncertainty: Public evidence indicates that stakeholders directly shaped AI design, data, deployment, evaluation, or oversight; residual uncertainty concerns the exact degree of shared decision power rather than project existence.