Meta policy shift linked to surge in online abuse against US lawmakers
The Center for Countering Digital Hate analysis of 8 million Facebook comments reveals a direct correlation between reduced enforcement and increased hate speech, drawing sharp criticism from politicians and experts.
Research conducted by the Center for Countering Digital Hate (CCDH) indicates that abusive and racist comments targeting US lawmakers tripled on Facebook within six months of Meta relaxing its content policies. The study, which analysed approximately 8 million comments, found that violent threats and hate speech quadrupled, while bullying and harassment doubled. Meta reduced its proactive content moderation enforcement by roughly half during this period. The report cites specific examples of gendered and racist abuse directed at representatives Jasmine Crockett and Byron Daniels, as well as a doubling of threats against President Donald Trump.
The investigation focused on 100 members of the House of Representatives, comprising 50 Republicans and 50 Democrats with the highest follower counts on the platform. Researchers utilised an AI system trained to identify comments likely to violate Meta’s current policies regarding violence, hateful conduct, and bullying. The data revealed that violent threats rose from 1,800 to 7,600, hate speech increased from 6,900 to 30,000, and bullying and harassment rose from 15,700 to 39,900 in the six months following the policy overhaul.
Meta’s chief global affairs officer, Joel Kaplan, previously stated the company had been "over-enforcing" rules and limiting legitimate political debate. However, the CCDH report argues that the surge in abuse and the collapse in enforcement track one another almost exactly. Meta’s own transparency reports from 2025 corroborate the reduction in proactive moderation, while a company spokesperson told WIRED that the prevalence of hateful conduct did not increase throughout the year, though the company declined to comment on specific examples provided by researchers.
Hours before the report’s publication, many of the abusive comments cited in the study were deleted from Facebook. Senator John Curtis, a Republican from Utah, stated that reducing oversight in areas like violence and hate should not be surprising given the increase in harms. Nina Jankowicz, CEO of the American Sunlight Project, argued that threats and abuse perform well on social media, keeping users engaged and driving ad revenue, thereby incentivising platforms to divest from content moderation.
The real-world implications of this digital environment are already being felt. Capitol Police cited increased threats to politicians in March when seeking a budget increase, and former US congresswoman Marjorie Taylor Greene referenced threats to her life as a reason for stepping away from public life. Imran Ahmed, CEO of CCDH, cited a "horrifying trend of political violence," referencing attacks on President Trump, the murder of Charlie Kirk, and the assassination of Melissa Hortman and her husband, warning that amplifying such content cannot be portrayed as a moral act.
