QueerTech report flags gaps in inclusive AI in Canada

QueerTech has published a report on inclusive AI development in Canada's technology sector, examining how companies account for 2SLGBTQI+ considerations in AI systems.

The study presents what QueerTech describes as a benchmark for responsible AI development and deployment in Canada. It draws on quantitative surveys and qualitative interviews with technology companies across the country.

Its findings point to a gap between stated commitments to inclusive AI and the processes, support, and knowledge in place within organisations. While 97% of companies surveyed said inclusive AI was a moderate or high priority, 43% reported limited or no formal processes for addressing bias in their work, and 71% reported limited or no organisational support for equitable representation.

The research also found a disparity between how companies viewed their products' suitability for the general population and for 2SLGBTQI+ users. Fewer than half of Canadian technology companies believed their AI products met the needs of 2SLGBTQI+ users, compared with 65% that believed their products met the needs of the general population.

Knowledge gaps

Another divide emerged between confidence and practical understanding. Although 85% of companies reported giving significant consideration to inclusion during the design, training, testing, and monitoring stages of AI product development, 77% said addressing queer inclusion across those same stages was challenging or very challenging.

The report also found that 18% of Canadian AI developers believed they had never encountered a single 2SLGBTQI+ safety consideration in their work. QueerTech said this reflected weak awareness of how identity-based risks can emerge in system design and deployment.

Companies also identified operational constraints. The main barriers to more inclusive development were insufficient resources or budget, cited by 39% of respondents, competing priorities at 36%, and difficulty measuring return on investment at 33%.

The study, funded by Women and Gender Equality Canada, was designed to capture views from a broader cross-section of the technology industry than QueerTech's earlier work. That approach was intended to compare awareness of exclusionary technology systems among queer and non-queer professionals.

Discriminatory responses

The research process itself also exposed hostility within the sector. QueerTech said 10% of responses submitted by Canadian technology companies during the project were homophobic, transphobic, or otherwise queerphobic.

Those responses did not come from bots, but from leaders working in Canada's technology industry. QueerTech argued that this illustrated the burden on groups researching underrepresented communities, which must also manage community protection and data verification.

Naoufel Testaouni, co-founder and chief executive officer of QueerTech, said the findings should inform both commercial decision-making and public policy.

“Technology serves whoever builds it. This study from QueerTech is a critical step in ensuring our most powerful technologies reflect and serve all Canadians,” said Naoufel Testaouni, co-founder and chief executive officer of QueerTech. “This report has established a new benchmark into how Canada is fairing in the era of AI; offering industry and government necessary insights to co-design informed regulatory and governance frameworks, build stronger products, and reinforce economic competitiveness through sustainable, strategic innovation.”

The report comes as governments and businesses face growing scrutiny over how AI systems affect access, safety, and outcomes for different groups. In Canada, the study frames those concerns through the experiences and needs of 2SLGBTQI+ communities.

The findings suggest that responsibility for inclusive development is often decentralised, while support structures are fragmented or missing. They also indicate that low awareness of inclusive development practices could lead to products that create identity-based disparities in access, usefulness, and risk.

Another broad theme was the contrast between public claims and internal readiness. Companies frequently reported that inclusion mattered to their AI work, yet many also acknowledged weak processes, low support, and significant difficulty applying inclusion in practice.

That tension is likely to sharpen debate over whether voluntary corporate efforts are keeping pace with the wider adoption of AI tools in products, services, and workplace functions. The report argues that legal, institutional, technical, and educational conditions will shape whether responsible and inclusive development becomes standard practice.

The findings also drew a response from the federal government.

“Artificial intelligence will help define Canada's economic future-but it must be shaped by all Canadians, not just a narrow slice of the tech sector. Organisations like QueerTech are leading the way, showing that when we build inclusively, we build smarter, fairer, and more competitive technologies,” said Solomon.