The week following the outbreak of the US-Iran war, an AI-generated video of Keir Starmer circulated on social media, accumulating approximately 400,000 views on Facebook and Instagram.
It depicts a believable likeness of the prime minister standing at a Downing Street lectern in Islamic dress. “We are still a proud Muslim country,” the avatar says. “Our crazy violent Muslim immigrants, including our precious criminals and rape gangs, are the bedrock of our society.” Later in the video, the avatar uses a racist term for Pakistani people.
AI-generated videos spreading misinformation and Islamophobia have already become common, especially following major news events. But this particular video was unusual. It had been posted by a devout Muslim man living in Pakistan.
This man, whom we are not naming for his own safety, runs successful accounts pushing Islamophobic AI content to UK audiences. Videos shared by just one of his Facebook pages had been viewed millions of times, and he told us in an interview he earned around $1,500 a month from that page alone.
He runs other accounts dedicated to sharing Quran verses and Islamic teachings. One features the Kaaba holy site in Mecca as its profile picture. He uses the title “Hafiz” before his name, an honorific reserved for those who have memorized the entire Quran.
Putin accused of surrounding himself with same 'actors' at series of events
We have covered the AI content industry before, revealing a Sri Lankan influencer pushing a similar kind of sensationalized racist content to UK audiences. But this case shows an even starker level of detachment between the creator and his posts: a Pakistani Muslim spreading content that portrays his religion as an existential threat to the West.
When we contacted him, he spoke candidly about his work, insisting he hadn’t understood the videos due to his poor English. He then deleted many of the posts.
“You are aware of the conditions in Pakistan, how petrol and circumstances are. Whoever is doing this work is doing it to make an earning,” he told us in Urdu. “We have no interest in news. I haven’t even looked at what is being said in the videos, what has been written and what hasn’t been written.”
The extent to which he truly grasped the nature of the content is unclear; it seems unlikely he could have missed the many visual clues. What’s certain is that viral content can now be produced so easily using AI tools that its creators don’t have to speak English or even fully understand what they are posting.
And at the heart of the issue is the way Meta has incentivized the creation of hateful AI content as it chases user engagement and ad revenue – while British Muslims suffer the consequences.
“This appalling example is yet more evidence of an ‘outrage economy’ where people are profiting from the poisonous narratives they push online, including those targeting Muslim Londoners,” said a spokesperson for the London mayor Sadiq Khan, who is a regular target in the man’s videos.
“Social media firms must do far more to stop the spread of lies and hatred on their platforms and prevent those creating and disseminating them from being financially rewarded.”
Accumulating views
The Pakistani creator ran a Facebook page called Britain Today, which had 192,000 followers, as well as an Instagram account with the same name which had 44,000 followers. Meta deleted both accounts after we flagged them, but not before their racist posts had been viewed millions of times. A Britain Today TikTok account, which had 11,000 followers, has also been deleted.
The Meta accounts have posted memes calling for all Muslims to be deported from the UK, propagating the “great replacement” conspiracy theory (which claims white populations are being deliberately replaced by non-white immigrants), and describing Muslims praying in public as a “dominance strategy” and an “invasion of the West”. Many of its Facebook videos received hundreds of thousands of views.
Edinburgh Hogmanay revellers stuck in queues for TWO HOURS in torrential rain
A spokesperson for Meta told us: “We have clear community standards that prohibit hate speech, harassment, harmful misinformation, and inauthentic behavior and we have removed these accounts for violating our policies.”
In a remarkably frank interview, the creator gave us a detailed insight into his day-to-day work. “All of the content, even the name of the page and everything, the content that is in the videos, that is AI,” he said. He mentioned using AI tools including Grok or Google’s image generator Whisk.
Sometimes, he said, he simply repurposed content from other platforms. “If there are protests and things in the UK, those videos I pick from Twitter or TikTok,” he said. “All of these things are copy-paste ... I lift these from there and put them on Facebook.”
When asked how he had learned these skills, he said he had paid other creators to learn monetization strategies and watched YouTube tutorials, going as far as to send us an example.
In the video, the creator opens Google’s AI chatbot Gemini, searching for trending news topics. After finding a summary of the latest on the US-Iran war, the user pastes the text into CapCut, a video editing tool owned by TikTok parent ByteDance. CapCut uses AI to generate a video based on this text, and the creator uses ChatGPT to create a thumbnail image.
He said he uses Meta’s content monetization tool, which enables creators to earn money from ad placements and bonuses on high-performing content.
Sam Stockwell, a senior researcher at the Centre for Emerging Technology and Security, described how these tools are changing the online landscape. “Beyond the use of AI for political disinformation, we are seeing the emergence of a ‘shadow influencer’ economy,” he said.
“This model prioritizes passive income over political ideology, turning divisive content into a profitable commodity. Exploiting the way social media algorithms prioritize high-engagement metrics, these creators are realizing that xenophobic or anti-establishment narratives are the most efficient pathways to virality – and therefore money.”
While many of Britain Today’s posts use AI-generated images, others take real footage out of context or add offensive captions. One post uses a real video in which Sadiq Khan speaks about prejudice faced by UK Muslims but adds a caption criticizing the government for funding a Muslim charity “while [Muslims] rape women and children.” (Data on the religious background of perpetrators of sexual assault is not routinely recorded and there is no evidence that Muslims are more likely to be perpetrators than other groups.)
Another real video reshared by the page showed Khan at a public iftar in Trafalgar square, described in a caption as a “colonization event.” The conspiracy theory that Muslims are “colonizing” the UK is a common trope in its posts.
Khan recently cited our previous investigation in a keynote speech at the Cambridge Disinformation Summit, calling it “shocking” that a man in Sri Lanka was profiting from “AI-generated content designed to stir up racial hatred in the UK.”
Emily Darlington, a Labour MP who sits on the Science, Innovation, and Technology select committee, said: “There’s clearly a market for hate content in the UK. When social media, intended to connect us, instead feeds us an endless stream of divisive and anger-fueling content it distorts not only how we feel about our neighbors but how we think the nation feels about our neighbors.
“The fact that this tactic is successful enough with a UK audience that individuals in other countries can profit off it shows how vulnerable we are. There’s nothing stopping foreign states from doing the same.”
‘What’s done is done’
When one of our reporters, a British Muslim, challenged the creator about the nature of his posts and the harm they caused to Muslims in the UK, he pleaded ignorance.
“I don’t speak proper English and then I don’t understand what they have and haven’t written,” he said in Urdu.
“But now what’s done is done. It’s a good thing you’ve told me. I am thankful to you. All of the posts that I have on the Facebook page, I’ll delete all of them … I now know that the immigration posts are against Muslims, so I’ll abstain from this now.
“I just knew that I was getting views and what else do I want? I did very wrong.”
However, many offensive posts, including the video of Starmer that used the P-word, remained live on Britain Today’s Instagram account until it was taken down by Meta.
“What this particular case shows is how generative AI tools now fundamentally industrialize [the creation of divisive content],” said Stockwell. “The barrier to entry has collapsed – a creator no longer needs to be a graphic designer or even fluent in the language of the target country.
“They simply type in a prompt request on an AI generator and the rest is taken care of. Coupled with a burgeoning ecosystem of online tutorials teaching others how to monetize this ‘outrage-for-profit’ model, it’s unlikely to be a problem that will be disappearing anytime soon.”
The Pakistani creator told us he has students who he teaches Facebook monetization. While others, including the Sri Lankan influencer from our previous investigation, charge for these tutorials, he claims to do this for free.
“Whatever knowledge I have, I share everything,” he said. “I don’t have any demand, I don’t have any greed, I don’t have any hopes for a reward … my only interest is that this person starts to make an earning.”
Deputy Editor & UK News
Read more similar news:
Comments:
comments powered by Disqus
