Nature

X's 2025 Transparency Report reveals a sharp decline to 2,326 hateful conduct suspensions in H2 2024, emphasizing 'freedom of speech, not reach' with focus on labeling over removal. A February 2026 Nature study confirms X's 'For You' algorithm amplifies conservative content, shifting user political attitudes rightward by 0.12 standard deviations. Meta's Q3 2025 report shows >90% enforcement precision and reduced errors, without ideological bias admissions; recent complaints of censorship appear more prevalent on other platforms like YouTube and TikTok.

Since early February 2026, new developments include Elsevier's Heliyon retracting 37 additional papers in 2026 amid an ongoing audit uncovering citation stacking and peer review compromise, following Clarivate's hold since 2024. ASTM's Journal of Testing and Evaluation reached 166 retractions by February 12 with enhanced safeguards against compromised peer review. Wolters Kluwer flagged 126 papers in Medicine on February 20 due to suspicious protocols. The International Journal of Obesity retracted a GLP-1 weight loss study on February 23 after statistical review found unsubstantiated claims. These updates affirm improved detection via PubPeer, audits, and editorial scrutiny, with no evidence of escalating fraud.