United News of Bangladesh

Meta's latest Integrity Report shows worrying spike in violent posts and harassment after policy shift aimed at easing restrictions on political expression. Facebook has seen a notable rise in violent content and online harassment following Meta’s decision to ease enforcement of its content moderation policies, according to the company’s latest Integrity Report. The report, the first since Meta overhauled its moderation strategy in January 2025, reveals that the rollback of stricter content rules has coincided with a drop in content removals and enforcement actions — and a spike in harmful material on its platforms, including Instagram and Threads. Meta’s shift, spearheaded by CEO Mark Zuckerberg, was aimed at reducing moderation errors and giving more space for political discourse. However, the company now faces growing concern that the relaxed rules may have compromised user safety and platform integrity. Violent Content and Harassment on the Rise The report shows that violent and graphic content on Facebook increased from 0.06–0.07 per cent in late 2024 to 0.09 per cent in the first quarter of 2025. While the percentages appear small, the scale is significant for a platform used by billions. Likewise, bullying and harassment rates rose in the same period. Meta attributed this to a March spike in violating content, noting a slight rise from 0.06–0.07 per cent to 0.07–0.08 per cent. These increases mark a reversal of a downward trend in harmful content seen in previous years. Content Removals and Enforcement Plummet The rise in harmful posts comes as Meta dramatically reduces enforcement activity. Only 3.4 million pieces of content were actioned under its hate speech policy in Q1 2025 — the lowest since 2018. Spam removals also fell sharply, from 730 million at the end of 2024 to 366 million in early 2025. Additionally, the number of fake accounts removed dropped from 1.4 billion to 1 billion. Meta’s new enforcement strategy focuses primarily on the most severe violations, such as child exploitation and terrorism, while areas previously subject to stricter moderation — including immigration, gender identity, and race — are now framed as protected political expression. The definition of hate speech has also been narrowed. Under the revised rules, only direct attacks and dehumanising language are flagged. Content previously flagged for expressing contempt or exclusion is now permitted. Shift in Fact-Checking Strategy In another major change, Meta has scrapped its third-party fact-checking partnerships in the United States, replacing them with a crowd-sourced system known as Community Notes. The system, now active across Facebook, Instagram, Threads, and even Reels, relies on users to flag and annotate questionable content. While Meta has yet to release usage data for the new system, critics warn that such an approach could be vulnerable to manipulation and bias in the absence of independent editorial oversight. Fewer Errors, Says Meta Despite the concerns, Meta is presenting the new moderation approach as a success in terms of reducing errors. The company claims moderation mistakes in the United States dropped by 50 per cent between the final quarter of 2024 and Q1 2025. However, it has not disclosed how this figure was calculated. Meta says future reports will include more transparency on error metrics. “We are working to strike the right balance between overreach and under-enforcement,” the report states. Teen Protections Remain in Place One area where Meta has not scaled back enforcement is in content directed at teenagers. The company has maintained strict protections against bullying and harmful content for younger users and is introducing dedicated Teen Accounts across its platforms to improve content filtering. Meta also highlighted growing use of artificial intelligence, including large language models, in its moderation systems. These tools are now exceeding human performance in some cases and can automatically remove posts from review queues if no violation is detected. As Meta pushes ahead with its looser content policies, experts and users alike will be watching closely to see whether the company can truly balance free expression with safety — or whether its platforms risk becoming breeding grounds for harmful content. Source: With inputs from agencies