Survey Finds Bias Gaps In AI Industry’s Efforts To Build Inclusive Products For LGBTQ+ Users – AfroTech
AI touches nearly every aspect of daily life, from features in products by companies like Apple and Google to influencing decisions in sectors such as hiring, healthcare, education, lending, and public services.
QueerTech, a Canadian-based nonprofit dedicated to helping LGBTQ+ individuals thrive in the technology industry, surveyed 100 AI product developers across Canada in December 2025. Respondents worked in leadership, engineering, product, operations, and responsible AI roles. Per the research, developers are still not building AI products with LGBTQ+ users in mind.
According to BetaKit, QueerTech co-founder and CEO Naoufel Testaouni said the survey responses reflect a split between unintentional gaps in understanding and more concerning attitudes towards the LGBTQ+ community held by some developers.
“Those are the people who are building these technologies,” Testaouni said, BetaKit reports. “If this is what they are thinking, and saying it publicly on a survey, then imagine all the other people also around them who are building this that we don’t even hear from.”
What QueerTech’s Survey Revealed About Inclusive AI
QueerTech found that fewer than half of respondents believe their AI products adequately serve LGBTQ+ users, compared with 65% who said their products meet the needs of the general population, per BetaKit.
The survey found that 97% of respondents considered building inclusive, bias-aware AI a moderate to high priority within their organizations. In comparison, 94% said formal processes are in place to identify and address potential bias. Developers also reported considering diverse representation throughout the AI product lifecycle, including during ideation and design (93%), model testing (91%), and ongoing monitoring (91%).
However, respondents indicated that progress in several key areas of inclusive AI development remains inconsistent. About 79% rated their use of diverse training data and ability to test for stereotypical outputs as good or excellent, while 70% said the same about representation of gender-diverse identities.
Despite widespread commitments to inclusive AI, only 29% of developers said their organizations provide full support for ensuring equitable representation in AI systems. Respondents cited limited resources, competing business priorities, and challenges demonstrating return on investment as the most significant barriers.
While AI has the potential to expand access and improve outcomes, QueerTech warned that AI systems can also reinforce existing biases if inclusivity is not prioritized throughout development and deployment.
Improving Responsible AI And Bias Practices
The organization emphasized that building inclusive AI extends beyond technical considerations and requires leadership commitment, workforce diversity, and accountability measures to maintain public trust. According to QueerTech, the responsible development of AI systems is key to protecting communities, driving meaningful innovation, and supporting long-term economic growth.
“Inclusive Al shouldn’t be optional. The systems being designed and deployed today are already shaping who gets hired, who gets funded, who is flagged, who is believed, and who is left out,” per the report. “As Al becomes more embedded in Canadian workplaces, products, and public services, the cost of getting it wrong grows exponentially. Bias, opacity, and uneven accountability do not remain contained within an algorithm; they travel outward into policies, customer experiences, reputations, and real lives. If we do not act with intention now, we risk building a future where inequity is automated, scaled, and normalized.”
As Pride Month continues and concerns about AI bias, responsible development, and inclusive design grow within LGBTQ+ communities, advocates say it is increasingly important to ensure AI systems meet the needs of diverse users.
