Anti-immigration AI videos traced to overseas fakers, BBC finds

Marianna Spring, Social media investigations correspondent

The Facebook page "Great British People," which claims to be based in Yorkshire, has garnered 1.3 million views for a video featuring an elderly white British man expressing sorrow over his pension. Additionally, other videos depict journalists discussing “the overwhelming scale of mass immigration” and question viewers if they miss “the Britain we used to know.”

However, it remains uncertain whether the individual producing these videos has any genuine connection to the UK, as the account is actually managed by someone located in Sri Lanka.

This page is among numerous interconnected accounts identified by BBC Panorama and the Top Comment podcast that generate and share anti-immigration AI-generated content concerning the UK for sizable audiences. The creators often operate from locations that are hundreds or thousands of miles away.

Many of these accounts originate from Sri Lanka, the US, and other European countries, while others are based in Vietnam and the Maldives, or linked to nations such as Iran and the UAE. This information has been gleaned from Facebook's transparency tools, conversations with content creators, and specific identifiers on social media, including language usage and the types of accounts they follow.

An expert informed the BBC that studies indicate that people are less capable of detecting AI-generated fakes than they believe, and exposure to more AI content can lead to increased distrust in genuine material.

London Mayor Sir Sadiq Khan, who has initiated research into AI-generated images depicting a declining capital, claims these misrepresentations damage London's international reputation. He articulated that while some individuals operating these accounts may be purely profit-driven, others appear to be supported by hostile states, such as Russia and Iran.

Although confirming direct state involvement poses challenges, a number of these accounts have shared posts that align with the narratives of the Russian and Iranian governments. Attempts to contact the account owners for clarification went unanswered.

Several accounts have altered their focus over time, seemingly to boost user engagement by transitioning from themes like "Make America Great Again" and "Life in the USA" to utilizing AI-driven narratives that undermine immigration perspectives. Some accounts have even occasionally showcased content that is more welcoming towards migrants.

According to Professor Sander van der Linden, a social psychologist at the University of Cambridge, these types of AI-generated accounts are a means through which states and organizations attempt to sway public opinion. He described them as a "new evolution of influence operations."

He pointed out that it is relatively inexpensive for overseas AI impersonators to masquerade as British nationals online by purchasing social media accounts that were originally created in the UK.

These accounts have accumulated vast audiences with AI-generated videos portraying fabricated scenes, such as a House of Commons filled with individuals in traditional Arab attire enforcing Sharia law, or fake interviews with women in hijabs contending that the UK requires greater Islamic influence.

The portrayal of the UK within these videos can be inconsistent. In some widely circulated content, societal decay is attributed to Muslim immigration, while concurrently, several videos from the same authors depict Islamic nations as idyllic in contrast.

Interviews with two individuals claiming to operate an account with over 20 million views indicated that their content features AI-generated visuals illustrating what British cities may look like in 2050. Cities such as Liverpool, London, and Birmingham are depicted as polluted and disorderly, populated by individuals in traditional Islamic attire, surrounded by stalls labeled “Halal” and featuring what appears to be Arabic-script bunting, amidst scenes of chaos.

Foreign cities like New York, Washington DC, and various European capitals are also represented in a similarly negative light.

When confronted about the divisive implications of their content, the creators stated, "Our content has a clear purpose: we aim to inform people and voters about what we believe could happen in the coming decades if current social and cultural trends continue."

They asserted that their operations were conducted from a European country where "a sense of insecurity has become more noticeable," claiming that they portray countries like Iran in an idealized manner to "provoke thought and discussion about political and cultural evolution."

The creators denied being motivated by profit and claimed their account is not monetized. They mentioned they maintain contacts with "various politicians" who are supportive of their content, although they declined to specify any names.

Research conducted by London's City Hall noted a significant rise in similar social media posts over the past two years, identifying two primary motivators.

"You've got state actors," London Mayor Sir Sadiq Khan detailed to the BBC. He noted evidence of activity from Russian and Chinese operatives, as well as involvement from "extreme right-wing" proponents of the Make America Great Again movement in the US.

"Secondly, we've seen individuals and companies trying to monetize and profit from division."

While acknowledging the challenges faced by the city, he pointed out that these "AI-generated lies" adversely affect London's perception, deterring visitors, international students, and investors.

"My concern is that decent individuals begin to believe these falsehoods, this dystopian portrayal of London being in decline, suggesting we are a dangerous city devoid of law and order," he emphasized.

Sir Sadiq urged social media companies to enhance their efforts in combating misinformation through measures such as "amending their algorithms to ensure they aren't rewarding toxic and divisive content" and clearly labeling AI-generated material.

In response, Meta—the parent company of Facebook and Instagram—asserted its commitment to counteracting "coordinated inauthentic behavior," stating that "specialized global teams are continually working to identify and disrupt this type of activity."

"We will take action against any content or accounts that violate our Community Standards, which apply to all content, regardless of whether it is created by AI or by a person," a spokesperson remarked.

Further inquiries with individuals managing accounts that cross-promote or interact with the 2050-themed videos confirmed City Hall's hypothesis that profit-driven motives were prevalent.

"I primarily post to elicit reactions for engagement, which boosts my follower count and income," stated one individual, who receives payments through Instagram's monetization program based on ads viewed alongside their videos.

Another respondent explained that they coordinate with accounts "raising awareness regarding similar issues," yet asserted that their online engagement is "not politically motivated." The primary objective, they mentioned, is for other accounts to share their content for maximum visibility.

While some individuals overseeing accounts that share similar content, and actively engage with these "fake" British patriots, are based in the UK, others are located in countries such as India, Pakistan, Singapore, Australia, and New Zealand.

Professor van der Linden remarked that the expanding “disinformation-for-hire industry” involves "paid actors and influencers pretending to be ordinary citizens to manufacture support for a specific agenda," often utilizing AI-generated content and bots to attract attention.

Research indicates that the public exhibits a limited capacity to identify fakes, with a mere 55% accuracy rate, as noted by Professor Yvonne McDermott Rees from Queen's University Belfast, who has explored the effects of deepfakes on trust.

This research also indicates individuals typically overestimate their ability to detect misinformation, she noted.

Responses to the AI-generated videos imply that certain viewers are indeed being misled. For instance, one woman commented, "keep going," while another asserted, "Never Back Down. Stand Up and Do What You Have to Do," in reply to an AI-generated video displaying a protester calling to "Stop the Boats" from bringing in migrants.

This particular video was posted on the "Britain Today" Facebook page, which, according to Meta's transparency tools, is operated from Sri Lanka.

"The more people are exposed to AI content, the harder it becomes for them to differentiate between fact and fiction, thus increasing their likelihood of distrusting authentic content," Professor McDermott Rees observed.

"It should not be the sole responsibility of the average citizen to discern what is real and what is not."

However, Professor van der Linden cautioned that research continues to suggest that many individuals seem indifferent regarding whether content is AI-generated.

"As long as it aligns with their identity and world-view, they are often willing to endorse and share the content because it signifies agreement with a broader agenda," he noted.

As one commenter aptly put it on the "Great British People" Facebook page: "It’s probably AI, but the fact is that he is right about everything."