Provenance Record

Well-structured News article covering GLAAD's 2026 Social Media Safety Index. Platform scores, policy rollback details, and real-world impact are clearly presented. The 'What We Don't Know' section adds editorial transparency. Word count of 769 falls comfortably within the News range (400-1200). No AI self-reference. Tone is appropriately neutral and factual throughout.

source-0.html.gz (advocate.com): Verified. Platform scores (TikTok 56, Instagram 41, Facebook 40, Threads 39, YouTube 30, X 29), Sarah Kate Ellis quotes, 92%/77% survey figures, 1,000+ ALERT Desk incidents, 20% hate crime statistic, transgenderism terminology claim, hashtag blocking in Meta teen search, and demonetization/shadowbanning claims all confirmed. Note: The article attributes the 'Both Meta and YouTube have implemented...' Ellis quote as 'said in a statement' but the Advocate source attributes it as 'wrote in the report.' Minor phrasing discrepancy. source-1.html.gz (gayety.com): Verified. Platform scores, YouTube 11-point drop as steepest decline, AI moderation suppression of LGBTQ content, 1,000+ incidents and 20% hate crime figure confirmed. shadowbanning and demonetization terms NOT found in Gayety snapshot — confirmed in source-2 (NBC Palm Springs). source-2.html.gz (nbcpalmsprings.com): Verified. Jenni Olson quote (Fundamentally, this signals to anyone on the platform that it is OK to express hate against trans people) confirmed verbatim. Platform non-response to comment requests confirmed. Shadow banning and demonetization confirmed. source-3.html.gz (attitude.co.uk): Verified. Consider ways to update the terminology and assessing feasibility Meta quotes confirmed. Instagram/Facebook/Threads scores confirmed. source-4 (qnews.com.au): Snapshot failed with HTTP 403. Fallback: WebFetch performed. QNews article confirmed platform scores and general AI moderation suppression claim, but did NOT mention LGBTQ+ hashtag blocking in Meta teen accounts — that claim is actually sourced from The Advocate (source-0), not QNews. This is a minor citation misattribution. source-5.html.gz (mambaonline.com): Verified. Leading social media companies today do not meet basic best practices Ellis quote and systemic failures quote confirmed verbatim. Real-world harm framing confirmed.

All platform scores are correct and confirmed across multiple sources. The 11-point YouTube drop, X last place, and TikTok stability are verified. The Meta policy changes list (hateful conduct policy, fact-checking end, Messenger themes, DEI dismantling, 'transgenderism' terminology) is verified in source-0. The 92%/77% survey figures and 1,000+ ALERT Desk incidents are confirmed. One citation misattribution found: the article attributes the LGBTQ+ hashtag blocking in Meta teen search tools to QNews, but the QNews snapshot (403; verified via WebFetch) does not contain this claim. It appears in The Advocate (source-0) instead. The fact itself is true and verified; only the inline citation is wrong. This is a correctable misattribution.

Solid, well-sourced news report on a significant annual LGBTQ+ safety index. One minor citation misattribution (LGBTQ hashtag/teen account claim credited to QNews instead of The Advocate) is correctable. All core facts are verified. APPROVE_WITH_CORRECTIONS is appropriate.