AI Bias Fuels Anti-LGBTQ Hate, GLAAD Warns

Artificial intelligence systems are increasingly reinforcing discrimination and bias against LGBTQ individuals, according to a new report from LGBTQ nonprofit GLAAD. The '2026 Build for Everyone' report details how AI can worsen misinformation, discriminatory decision-making, and privacy concerns across the AI lifecycle.

GLAAD found that if LGBTQ topics are not accurately represented in AI development, systems can perpetuate stereotypes. For instance, a previous report identified issues with a Meta model repeating misinformation about conversion therapy when users inquired about suppressing same-sex attraction. AI-enabled censorship is also a concern, as automated content moderation systems often fail to understand LGBTQ nuances, leading to the targeting of queer content.

Furthermore, AI used in predictive decision-making, such as in banking, housing, and hiring, can amplify historical discrimination and flawed assumptions about identity groups. The report also emphasizes heightened data privacy risks for LGBTQ people, as AI-collected information can lead to severe consequences, including arrest or violence in countries criminalizing same-sex relationships, and discrimination or denial of care in jurisdictions restricting transgender rights.

GLAAD recommends that AI developers increase LGBTQ representation in training data to address model blind spots and continuously update models. They also call for intentional guardrails and stress testing of AI products with LGBTQ communities in mind. GLAAD president and CEO Sarah Kate Ellis stated that "neutrality is no longer an option" and that tech leaders must proactively create safe products, as responsible AI is crucial for business and future-proofing companies.