Policies Public-facing policies
AI Training and Data Use Policy
How Map.ca's user content is and is not used for training AI — Map.ca's own systems and external vendors alike.
- Version
- 0.1.0
- Effective
- May 20, 2026
- Last reviewed
- May 20, 2026
- Review cycle
- every 6 months
- Master Policy Index entry
- §4 #33
Policy text
Default: no. Map.ca user content is not used to train AI models, internal or external, without separate, explicit, plain-language permission. This policy defines what permission looks like (granular, revocable, separately collected from the general Terms acceptance), what user content categories are eligible for opt-in training, and what categories are never eligible (sensitive locations, content involving minors, Indigenous community data, civic reports involving people).
It applies to the Map.ca AI team, AI vendors, and any third party that touches Map.ca user content with a model-training pipeline.
Principles this policy enforces
- AI assists, people remain responsible
- Consent must be meaningful
- Collect less, protect more
- Public data and personal data are not the same thing
- Indigenous data sovereignty is foundational
What it requires
- Get separate, explicit, granular consent before using user content for AI training.
- Make the consent revocable and document the revocation effect.
What it forbids
- Do not train AI models on user content without explicit opt-in.
- Do not train on sensitive-location, minor, Indigenous, or civic-report content at all.
How it applies
- Map.ca AI team
- AI vendors
- Researchers using Map.ca content
References
- Map.ca Policy Constitution §2 principles 5 and 6
Related policies
AI Use Policy
Where AI is used on Map.ca, what it can and cannot do, and how humans stay responsible for outcomes.
AI Vendor Policy
Standards that AI vendors must meet to handle Map.ca traffic — data handling, training restrictions, audit requirements.
Indigenous Data Sovereignty Policy
How Map.ca handles Indigenous community data, cultural places, traditional knowledge, names, language, heritage, and map layers.