AI Bias Is Putting LGBTQIA+ People at Risk

2026-06-25 · Source: Partnership on AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy · Depth: Intermediate, short

Summary

GLAAD CEO Sarah Kate Ellis highlighted the risks of AI bias to LGBTQIA+ individuals at the Axios' AI+ NY Summit, coinciding with the release of GLAAD's "Build for Everyone: A Framework for LGBTQ Representation and Safety in AI" report on June 25, 2026. This report details how algorithmic discrimination, stemming from distorted training data, leads to unjust outcomes. For instance, a 2024 UNESCO study revealed Meta's Llama 2 model produced negative content about gay people in approximately 70% of instances, characterizing them as criminals or abnormal. Beyond biased outputs, AI systems can infer LGBTQIA+ identity from user data, enabling surveillance, as documented by Human Rights Watch regarding government targeting. The challenge is compounded by a hostile political environment, with 797 anti-LGBTQIA+ bills filed in the U.S. in 2026. Addressing bias requires demographic data, yet its collection poses risks like miscategorization and inadequate security, which PAI's Participatory & Inclusive Demographic Data Guidelines aim to mitigate.

Key takeaway

For Directors of AI/ML developing systems that process sensitive user data, you must prioritize robust privacy-by-design principles and integrate community engagement from the outset. Your teams should adopt PAI's Participatory & Inclusive Demographic Data Guidelines to assess fairness without exposing vulnerable groups. Failing to implement these safeguards risks severe algorithmic discrimination and erosion of trust, especially for LGBTQIA+ communities facing increasing digital targeting and legislative hostility.

Key insights

AI bias disproportionately harms LGBTQIA+ communities, necessitating responsible demographic data practices and community-led safeguards.

Principles

• Algorithmic fairness requires demographic data.
• Data collection can itself be a source of harm.
• Affected communities must shape protective standards.

Method

PAI's Participatory & Inclusive Demographic Data Guidelines offer a framework for organizations to collect and use sensitive demographic data for fairness assessments responsibly.

In practice

• Adopt privacy-by-design principles.
• Engage civil society experts early in AI development.
• Utilize PAI's Demographic Data Guidelines.

Topics

• AI Bias
• LGBTQIA+ Rights
• Algorithmic Discrimination
• Data Privacy
• Demographic Data Guidelines
• AI Ethics

Best for: CTO, VP of Engineering/Data, Executive, AI Ethicist, Policy Maker, Director of AI/ML

Related on AIssential

• An Industry Benchmark for Data Fairness: Sony’s Alice Xiang — Ethically sourced, diverse datasets are crucial for effectively measuring and mitigating bias in AI models.
• » Detroiters cautious about government use of AI, U-M survey finds - Poverty Solutions — Public support for government AI applications is conditional, varying by use case and significantly impacted by data privacy concerns.
• Watch: Artifical intelligence and algorithms perpetuate biased housing - AOL.com — AI and algorithms may perpetuate biased housing decisions.
• The Top 10 PAI Blogs and Resources of 2025 — AI discussions in 2025 focused on ethical implications, societal impact, and practical applications.
• AI is exacerbating gender imbalances in tech and finance — AI is worsening gender disparities in tech and finance through job displacement, adoption gaps, and embedded biases.
• Promise and Perils of Using AI for Hiring: Guard Against Data Bias — AI in hiring offers efficiency but demands careful implementation to prevent discrimination.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Partnership on AI.