Scroll through any Facebook feed in Britain, and between baby announcements and neighborhood disputes, you are likely to encounter accounts with Union Jack profile pictures and generic names like Britain Today. These pages, numbering in the hundreds or thousands, present themselves as patriotic British voices. One typical AI-generated video shows a middle-aged man claiming his local cafe has stopped serving pork to avoid offending people. Another post features sepia-tinted images of Victorian London, lamenting a time when the city was English and first-world. Alongside this nostalgic content, memes call Islam a cancer, decry Muslims praying in public as an invasion of the West, or promote the great replacement theory.
Who Is Behind These Pages?
For the past seven months, the Bureau of Investigative Journalism has investigated the origins of these pages. The answer often points to young, entrepreneurial men in South Asia. They have little interest in UK politics but create content that boosts far-right talking points and fuels hostility toward immigrants and British Muslims. They are part of a booming cottage industry producing commercial AI slop.
Financial Incentives
The financial rewards are huge, especially for creators in the global south. The Bureau examined two successful sloperations targeting British audiences from Pakistan and Sri Lanka. These creators earn money from ads placed by Meta next to high-performing content, with Meta sharing ad revenue and making direct payments for engagement. A Pakistani creator, a devout Muslim who remains unnamed for safety, earns $1,500 monthly from one page. Sri Lankan creator Geeth Sooriyapura claims to have made $300,000 over his Facebook career. While unverifiable, these sums far exceed average incomes in their countries. Their success promotes passive income culture, with Sooriyapura claiming 2,500 graduates from his content academy.
Structural Factors Enabling Hate Content
Two key factors make this content pervasive. First, generative AI tools are used to brainstorm ideas, write captions, and create compelling images and videos. This helps creators who lack English fluency. Sooriyapura told students that AI-generated videos can make political content go viral ten times faster. Second, Meta has retreated from content moderation, laying off trust and safety teams amid pressure from the Trump administration. Meta claims AI can efficiently find harmful content, but our reporting shows deeply offensive content remains easily accessible.
Meta's Response
After we contacted the Pakistani creator, he deleted many posts, calling it a good thing. Sooriyapura denied encouraging violence, stating he only teaches monetization and audience targeting. Meta took down many pages after our inquiry, stating they violate policies against hate speech and harassment. However, the Sri Lanka network is back online with minimal consequences. Meta could do more, but as long as its algorithm rewards extreme emotions, such content will persist.
