GLAAD, an influential LGBTQ media advocacy organization, released a report titled “Build for Everyone: A Framework for LGBTQ Representation and Safety in AI” in June 2026. The report outlines the pervasive biases embedded in artificial intelligence systems that negatively affect LGBTQ individuals and offers a comprehensive framework for creating safer, more inclusive AI technologies.
What Happened
On June 2026, GLAAD published “Build for Everyone: A Framework for LGBTQ Representation and Safety in AI,” drawing upon years of research and collaboration with other civil rights groups, including the Leadership Conference on Civil and Human Rights and UNESCO. The report pinpoints key areas where AI perpetuates harm through biased training data, misinformation, and harmful stereotyping specifically targeting LGBTQ identities.
GLAAD’s senior director of the Social Media Safety Program, Jenni Olson, discussed how the report builds on prior work dating as far back as 2018 with AI companies such as Alphabet, focusing on eliminating LGBTQ slurs and bias. However, the rapid adoption of generative AI models in recent years has accelerated the urgency to address these harms systematically.
Key Facts
The report highlights that many AI language models, including Meta’s Llama 4, have displayed bias by promoting conversion therapy content despite existing policies prohibiting such promotion. Conversion therapy, widely condemned by medical associations and the United Nations as pseudoscientific and abusive, was still promoted by AI under specific user prompts, illustrating the dangerous misinformation AI can propagate.
GLAAD also references Meta’s policy shifts in January 2025, which introduced explicit anti-LGBT rhetoric into its community guidelines, signaling deeper ideological conflicts within AI governance and platform management. The report stresses that bias in AI models arises both from technical limitations in training data and from political or corporate choices about product deployment and content moderation.
Crucially, the report calls out the lack of transparency around AI algorithms and moderation systems, noting that communities affected by AI harms have limited access to recourse or accountability from companies deploying these systems. This opacity complicates efforts to correct or mitigate discriminatory practices.
What This Means
GLAAD’s report reveals the real-world consequences of AI bias on LGBTQ people, including exposure to harmful stereotypes, misinformation, and dangerous content like conversion therapy promotion. This not only undermines the dignity and safety of LGBTQ individuals but also erodes trust in AI technologies and the platforms that use them.
By framing these issues as both technical and political, the report emphasizes that addressing AI bias requires intentional design decisions, transparent policies, and a commitment from AI developers and tech companies to prioritize safety and representation. For policymakers and regulators, this report underscores the necessity of establishing clear standards and oversight mechanisms for AI systems to protect marginalized communities.
For the general public, the findings highlight the importance of demanding ethical AI that respects diversity and human rights. The report challenges the notion that AI bias is inevitable, presenting it instead as the product of choices—both in data curation and governance—thus opening pathways for meaningful reform.
Background
GLAAD’s involvement with AI issues began in 2018, with collaborative efforts targeting the reduction of LGBTQ-related slurs in AI outputs. Over the past several years, GLAAD’s work extended to broader AI safety concerns through its Social Media Safety Program and research initiatives.
The organization also builds on the frameworks developed by the Leadership Conference on Civil and Human Rights and others, who have addressed overlapping civil rights implications of AI bias impacting various marginalized groups. The report synthesizes these existing frameworks specifically through the lens of LGBTQ representation and safety.
What Remains Unclear
The report does not address the extent to which companies have implemented GLAAD’s recommended practices or how ongoing policy changes, such as Meta’s new community guidelines, will affect AI moderation moving forward. It also remains uncertain how regulators will integrate these advocacy-driven frameworks into concrete AI regulations or how enforcement might proceed in the absence of transparency from AI developers.
What Comes Next
GLAAD continues to engage directly with AI companies, advocating for adoption of best practices around safety, privacy, and inclusion. The organization emphasizes collaboration while maintaining pressure on firms to improve AI products responsibly. Scheduled legislative and regulatory developments around AI safety standards remain forthcoming, and the report signals that civil society will play a vigilant role in shaping these processes.
Sources
This article is based on reporting and publicly available information from the following source:
Read more AI Regulation stories on Goka World News.