GLAAD 2026 Social Media Safety Index Shows No Platform Passes LGBTQ+ Test

The 2026 GLAAD Social Media Safety Index shows that no platform earned a passing score for protecting LGBTQ+ users. All six evaluated platforms received failing marks on safety, privacy, and expression metrics.

The report, covering data through mid-2026, documents continued declines tied to specific policy retreats at Meta and YouTube while highlighting broader transparency shortfalls.

Overview of the 2026 GLAAD Social Media Safety Index

The central conclusion from the 2026 GLAAD Social Media Safety Index is that every major platform fails to provide adequate protection for LGBTQ+ communities through its policies and practices. This assessment rests on a structured review of 14 LGBTQ-specific indicators that examine content moderation, transparency, data privacy, and expression.

The mechanics of the index adapt the Ranking Digital Rights methodology to create targeted benchmarks for LGBTQ+ safety. Evaluators review public policy documents, transparency reports, and observed platform behaviors to assign points across each indicator, with a maximum total of 100 points. A passing score requires consistent performance that demonstrates explicit safeguards against targeted risks.

Criteria for scoring include whether policies explicitly prohibit anti-LGBTQ rhetoric, whether reporting tools are accessible and effective, and whether algorithms and AI systems receive documented oversight to prevent wrongful suppression. Platforms also receive credit for maintaining privacy standards that protect user data from misuse in content decisions.

Limitations of the index include its reliance on publicly available policy statements rather than internal enforcement statistics, as the report notes no direct access to moderation data. Scores represent conditions during the primary analysis period in early to mid-2026, and GLAAD as an advocacy organization applies its own priorities to the indicators. Subsequent platform updates after the report date may not appear in the results.

A conditional practical example would involve a platform that maintains detailed privacy policies but permits specific anti-LGBTQ terms in its conduct rules, resulting in point deductions on multiple indicators and an overall failing mark. This type of gap shows how isolated policy weaknesses can prevent a passing total even when other areas perform adequately.

Typical errors when reviewing the index include assuming that a higher score means complete safety for all users rather than targeted policy evaluation, or overlooking that the methodology focuses exclusively on LGBTQ+ indicators instead of broader platform performance. Another common mistake is treating the scores as permanent when they reflect a fixed analysis window and may shift with new disclosures.

2026 Platform Scorecard and Rankings

The 2026 platform scores place TikTok at the top with 56 points while X sits at the bottom with 29 points, and the remaining platforms fall in the middle range without any reaching a passing threshold. These results reflect uniform application of the 14 indicators across all six services.

The mechanics behind the scorecard involve breaking down each platform's policies into the 14 indicators and assigning points based on alignment with LGBTQ+ safety standards. TikTok maintained its prior score while the others declined, indicating that most platforms introduced or retained changes that reduced their performance on key metrics.

Criteria for the rankings include performance on hate speech definitions, content visibility rules, transparency around algorithmic amplification, and commitments to data handling practices. Higher scores require evidence that policies actively mitigate disproportionate impacts on LGBTQ+ expression and privacy.

Limitations include the fact that scores derive from policy analysis as of the report period and do not incorporate raw enforcement numbers from the platforms themselves. The index also notes that external factors such as regulatory changes after mid-2026 could alter future evaluations.

A conditional practical example would involve a platform that improved its transparency reporting on AI moderation, potentially raising its score by several points if the change addressed wrongful suppression of LGBTQ+ content. Such an adjustment would still need to meet all other indicators to approach a passing level.

Typical errors include misinterpreting year-over-year declines as temporary without checking the specific policy drivers, or assuming that the highest-ranked platform offers reliable safety when its score remains well below the passing mark. Readers may also overlook that the rankings compare platforms only within this LGBTQ+-focused framework.

Policy Changes Driving Score Declines

The conclusion is that Meta and YouTube policy adjustments from early 2025 directly lowered their 2026 scores by weakening protections against anti-LGBTQ rhetoric and reducing oversight mechanisms. These changes affected multiple indicators related to hate speech and content moderation.

The mechanics of the impact involve how specific wording in conduct policies influences point allocation. When platforms modify rules to allow previously restricted language or remove categories from protected characteristics, evaluators deduct points on the corresponding indicators for safety and expression.

Criteria for evaluating these changes include whether the updated policies explicitly permit terms such as 'transgenderism' or 'mentally ill' in certain contexts, and whether gender identity remains listed among protected characteristics in hate speech guidelines. Meta also ended U.S. fact-checking programs and scaled back DEI initiatives, which further reduced scores on transparency and accountability indicators.

Limitations arise because the report relies on public announcements of these changes rather than internal data showing enforcement frequency. The analysis covers the period up to mid-2026, and any reversals or additional modifications after the report publication date fall outside the evaluated scope.

A conditional practical example would involve a platform that reintroduces gender identity protections in its hate speech policy, which could restore points on the relevant indicator and improve the overall score if other areas remain stable. This would require clear documentation to receive credit in future assessments.

Typical errors include attributing score declines solely to one company without recognizing parallel changes across multiple platforms, or assuming that policy language alone determines real-world outcomes without considering how moderation teams apply the rules in practice.

Key Findings on Hate, Suppression, and Transparency

The main conclusion is that platforms disproportionately suppress LGBTQ+ content while failing to provide meaningful transparency on moderation decisions, algorithms, AI use, and data privacy practices. Five core issues surface from the evaluation of policy enforcement across the six platforms.

The mechanics of these findings involve comparing stated policies against observed patterns of content removal and visibility reduction. When platforms apply broad AI filters without sufficient human oversight, the result often includes erroneous suppression of legitimate LGBTQ+ material alongside inadequate handling of violating hate speech.

Criteria for identifying these problems include the presence of documented cases where content is removed without clear explanation, the absence of regular transparency reports on algorithmic amplification, and the rollback of diversity commitments that previously supported specialized moderation training. Over-reliance on AI contributes to inconsistent application of rules.

Limitations include the report's dependence on public disclosures and external incident data rather than comprehensive internal statistics. The findings reference patterns from 2025 and early 2026 but do not quantify exact volumes of suppressed content for each platform.

A conditional practical example would involve a platform that publishes detailed AI decision logs showing how content is flagged, allowing evaluators to verify whether suppression rates differ for LGBTQ+ topics. Without such logs, the platform loses points on transparency indicators.

Typical errors involve assuming that all content removals reflect accurate enforcement rather than overbroad AI application, or ignoring that lack of transparency prevents independent verification of whether policies are followed consistently.

Recommendations for Platforms

The report concludes that platforms should strengthen enforcement of existing hate speech policies, improve moderation training on LGBTQ+ issues, and increase transparency around algorithmic and AI decisions. These steps address the documented gaps in safety and accountability.

The mechanics of implementing recommendations involve updating policy language to restore explicit protections, reinstating fact-checking programs where applicable, and publishing regular reports on how AI systems handle content related to protected characteristics. Data privacy practices also require clearer user controls and disclosure.

Criteria for effective recommendations include measurable actions such as mandatory training modules for moderators, public dashboards on moderation outcomes, and commitments to maintain DEI programs that support specialized review teams. Platforms must also address over-reliance on automated systems by adding human review layers.

Limitations of these recommendations include the fact that they depend on voluntary platform adoption, as the index does not have enforcement authority. The suggestions reflect analysis up to mid-2026 and may require adjustment if new regulatory requirements emerge.

A conditional practical example would involve a platform that releases quarterly transparency reports detailing AI moderation accuracy rates for LGBTQ+ content, which could demonstrate progress on the transparency indicators in a subsequent index evaluation.

Typical errors include viewing recommendations as optional suggestions rather than necessary responses to failing scores, or implementing changes only in public statements without corresponding updates to actual moderation practices and training programs.

Context of Broader Online and Offline Harms

The conclusion is that online anti-LGBTQ hate and disinformation on platforms contribute to increased offline incidents, as shown by GLAAD ALERT Desk data recording more than 1,000 incidents in 2025 and FBI statistics indicating anti-LGBTQ bias in over 20 percent of hate crimes for the third consecutive year based on 2024 figures.

The mechanics of this connection involve how platform failures to mitigate violating content allow hate and disinformation to spread, which the report links to real-world bias-motivated events. Political developments since 2025 have amplified exposure when moderation does not align with stated policies.

Criteria for establishing the link include references to external data sources that track incident volumes and crime statistics, combined with platform policy analysis showing inadequate responses to hate speech that violates their own rules. The index emphasizes that suppression of legitimate content occurs alongside insufficient action on prohibited material.

Limitations include the report's use of aggregated external data without providing new primary platform enforcement statistics. The connections draw from cited surveys and reports but remain observational rather than establishing direct causation for every incident.

A conditional practical example would involve a platform that consistently removes hate speech violating its policies, potentially reducing the volume of content that contributes to offline risks if the enforcement is applied evenly across topics.

Typical errors include dismissing the offline connections as unrelated to platform practices, or assuming that all reported incidents stem solely from online activity without considering multiple contributing factors in the broader environment.

Implications for LGBTQ+ Community Users

The conclusion is that LGBTQ+ users face ongoing challenges in finding reliable online spaces due to the documented policy and transparency shortfalls across platforms. The index emphasizes the need for community features that operate without unwarranted suppression.

The mechanics of these implications involve how failing scores translate into reduced visibility for LGBTQ+ content and increased exposure to hate when platforms do not enforce their rules consistently. Users encounter these effects through algorithmic decisions and moderation outcomes that the report identifies as problematic.

Criteria for assessing user implications include survey responses on concerns about platform reliability and the availability of safe community spaces. The report highlights that maintaining expression without suppression supports community connection while inadequate transparency leaves users without clear recourse.

Limitations include the absence of new primary user data in the index itself, with impacts drawn from referenced external sources. Scores and findings reflect policy analysis rather than direct measurement of individual user experiences.

A conditional practical example would involve users cross-referencing platform scores when selecting where to share content, allowing them to prioritize services that perform better on the evaluated indicators while remaining aware of the overall failing results.

Typical errors include assuming that lower-ranked platforms are entirely unusable or that higher-ranked ones require no caution, without recognizing that all platforms in the index fall short of passing standards and require continued monitoring for policy updates.