New Research: Indigenous Workers, Training, and AI
The Diversity Institute at Toronto Metropolitan University has released Indigenous Peoples in the Workplace: Examining Employment, Discrimination, Training and Artificial Intelligence Use (2026), drawing on Wave 8 of the Survey on Employment and Skills, fielded across Canada in spring 2025.
Canada's Indigenous population is growing at nearly twice the rate of the non-Indigenous population but remains underrepresented in the labour market — a gap the report attributes to systemic barriers: a history of genocide, discrimination, exclusion, intergenerational trauma, and lack of access to infrastructure.
What the data shows:
Training pays off. Indigenous Peoples participate in workplace training at a higher rate than non-Indigenous workers — and are more likely to see raises, promotions, and moves into better jobs as a result.
AI adoption is ahead, not behind. Indigenous respondents are more likely to use AI tools at work, and more likely to use them for productivity and creativity. The digital divide framing doesn't hold here.
Discrimination remains widespread. About 38% of Indigenous respondents reported discrimination based on their Indigenous identity — rising to 45% among those aged 18 to 34.
Education pathways differ. 53% of Indigenous respondents reported a high school diploma or less (vs. 19% of non-Indigenous respondents), but they were more likely to have completed apprenticeship or trades training.
Satisfaction and outlook run high. Indigenous respondents reported higher job satisfaction, more optimism about their finances, and a more positive view of where Canada is heading — alongside poorer mental health outcomes.
The recommendations: reduce workplace discrimination, address education barriers with targeted supports for Indigenous youth (especially young women), and expand mentorship opportunities.
The picture that emerges: strong engagement, strong training uptake, and early AI adoption, running up against structural barriers that individual motivation can't fix.
