Table of Contents
AI has stoked many fears: The demise of entire industries, mass exploitation of vulnerable online denizens, a widespread crisis of literacy and critical thinking.
Included in that long list is the fear that these systems can reinforce discrimination and bias levied against already marginalized groups. And according to a new report from LGBTQ nonprofit GLAAD, it's already a problem.
The 2026 Build for Everyone report, published by GLAAD today (June 17), is an industry-wide analysis of the state of inclusive, responsible model design across the AI lifecycle. Researchers found repeated instances of exacerbated misinformation, discriminatory decision-making, and privacy concerns that the organization argues must be addressed with more responsible model design. But it has to happen sooner rather than later, GLAAD argues.
AI can make misinformation and censorship worse
"If LGBTQ topics and issues aren’t accurately represented during foundation model development or model fine-tuning, AI systems can perpetuate biased or stereotypical assumptions," GLAAD writes in its report.
For example, a 2025 GLAAD report found that Meta’s Llama 4 model repeated harmful information about conversion “therapy" — an anti-LGBTQ practice that has been disavowed by the majority of medical professionals and even the United Nations — when queried by users about how to "stop" same-sex attraction. Mainstream generative AI chatbots have been known to repeat medical misinformation, especially about politically charged topics like abortion.
Mashable Light Speed
In addition, GLAAD warns of AI-enabled censorship of LGBTQ content as social media platforms rely increasingly on automatic content moderation. These systems often fail to parse the nuances of queer identities or outrightly target LGBTQ content. Meta's Oversight Board, for example, has urged the company to improve automatic enforcement of its Hateful Conduct policy following a complete overhaul of its LGBTQ protections.
AI-based prediction can reinforce discrimination
GLAAD also found that systems relying on AI systems for predictive decision-making — like AI-powered banking or housing processes, job-hiring tools, and even ad targeting — can further exacerbate historically discriminatory practices, worsen stereotypes, and repeat flawed assumptions about identity groups.
AI poses data privacy concerns for LGBTQ people
Data privacy has been a prevailing issue among AI users and developers, and the same goes for LGBTQ communities.
"LGBTQ people face heightened risks when AI systems collect, infer, or retain information about sexual orientation, gender identity, or other personal characteristics," writes GLAAD. "In the more than 60 countries that criminalize same-sex relationships, government access to AI-collected information can lead to arrest, persecution, or violence. In the many U.S. jurisdictions that restrict transgender rights, that same data can fuel discrimination, denial of care, or loss of legal recognition."
How to design AI with LGBTQ communities in mind
GLAAD recommends AI developers ensure greater LGBTQ representation in AI training and data to fill in model "blind spots," and urges companies to continuously update their models as anti-LGBTQ hate and misinformation evolve.
GLAAD also recommends that companies include intentional guardrails to protect LGBTQ users and that AI products are stress tested and deployed with LGBTQ communities in mind.
"Neutrality is no longer an option. AI systems trained on data that wrongfully positions LGBTQ lives as 'fringe' or treats our equal rights as 'controversial,' or that fail to catch sophisticated disinformation about LGBTQ people, threaten our health, safety, and civil rights," GLAAD president and CEO Sarah Kate Ellis said in a press release. "Tech leaders must proactively embrace intentional practices to create safe products — not just because it is an urgent moral imperative, but because responsible AI is best for business and a requirement for future-proofing AI companies.”