Key facts
- GLAAD's report, "Build for Everyone: A Framework for LGBTQ Representation and Safety in AI," details how AI bias threatens LGBTQ users.
- The report argues that AI trained on biased data can reinforce stereotypes, suppress LGBTQ voices, and lead to discriminatory outcomes.
- Concerns include biased training data, anti-LGBTQ misinformation, discriminatory outcomes in predictive AI, content moderation failures, and privacy risks.
- GLAAD warns that autonomous AI agents could automate discrimination in areas like housing, employment, healthcare, and lending.
- The report calls for improved LGBTQ representation in training data, stronger privacy protections, human oversight, and industry accountability.
GLAAD has released a report titled "Build for Everyone: A Framework for LGBTQ Representation and Safety in AI," detailing how artificial intelligence systems are amplifying anti-LGBTQ bias, misinformation, and discrimination. The report argues that LGBTQ safety must be a core requirement of responsible AI development, as biased training data can reinforce stereotypes, suppress voices, and lead to discriminatory outcomes in areas such as housing, employment, healthcare, and lending. GLAAD President and CEO Sarah Kate Ellis stated that AI is a civil rights issue and that neutrality is no longer an option, emphasizing the need for tech leaders to proactively create safe products. The report highlights the significant economic power of the LGBTQ community, noting that over 20% of Gen Z identifies as LGBTQ and possesses substantial global buying power. The findings come amid broader debates on AI bias, including research showing favoritism towards Catholicism in AI models and a lawsuit against xAI alleging its Grok chatbot lacked adequate safeguards against misinformation and bias. GLAAD calls for improved LGBTQ representation in training data, stronger privacy protections, human oversight, and greater industry accountability to prevent these risks from becoming further embedded in AI systems.