A BBC investigation has uncovered a sprawling network of fake social media accounts, based overseas, that are using generative artificial intelligence to produce and spread anti-immigration videos targeting the United Kingdom. The findings reveal a sophisticated disinformation campaign designed to inflame tensions and manipulate public opinion ahead of the next general election.
The investigation identified dozens of accounts, primarily originating from Russia and Eastern Europe, that posted hundreds of AI-generated videos on platforms such as TikTok, X (formerly Twitter), and Facebook. These videos, often depicting lurid and baseless scenes of immigrants committing violent crimes or abusing public services, were crafted using advanced deepfake technology and text-to-video models like OpenAI's Sora and Stable Video Diffusion. Many were indistinguishable from genuine footage to the untrained eye.
One video, viewed over 2 million times, showed a fake news report about a Somali migrant attacking a police officer in Manchester. The BBC traced its upload to a server in St Petersburg, linked to a known Russian troll farm. Another, featuring a fabricated speech by a British MP calling for open borders, was generated using a voice cloning algorithm that mimicked the politician's cadence and accent with alarming accuracy.
Dr. Elena Petrova, a disinformation researcher at the University of Oxford, told the BBC: 'This is a new frontier in information warfare. AI tools have lowered the barrier to creating convincing fake content. We are seeing coordinated attempts to weaponise these technologies against democratic societies.'
The campaign appears to exploit existing societal divisions, particularly around immigration and national identity. The videos often target far-right and anti-immigration groups for amplification, seeding distrust in mainstream media and political institutions. Some have been shared by verified accounts with hundreds of thousands of followers.
In response, the UK government has announced a new task force, the Digital Integrity Unit, to combat AI-generated disinformation. A Downing Street spokesperson said: 'We will not tolerate foreign interference in our democracy. We are working with social media platforms to rapidly identify and remove such content, and we are investing in AI detection tools to stay ahead of these threats.'
But critics argue that the response has been too slow. 'The government has known about these capabilities for months,' said Professor Tim Harper, a cybersecurity expert at King's College London. 'They need to act with greater urgency. Every hour a fake video stays online, it causes real-world harm.'
The BBC investigation highlights the challenges of regulating AI-generated content. Current laws require platforms to remove illegal content, but deepfakes often fall into a legal grey area unless they incite violence or hatred. Social media companies have faced criticism for inconsistent enforcement of policies against synthetic media.
Meta, which owns Facebook and Instagram, said it had removed a 'small number' of accounts linked to the campaign and was developing automated tools to identify AI-generated content. TikTok said it had taken down several videos for violating its misinformation policies. X, meanwhile, has been under fire for reducing content moderation since Elon Musk's takeover in 2022.
The broader implications are unsettling. As generative AI becomes cheaper and more accessible, the volume of synthetic disinformation is likely to explode. Researchers warn that we may soon reach a 'tipping point' where it becomes impossible to distinguish reality from fabrication, eroding trust in all digital media.
For the average British citizen, the message is clear: approach viral videos with scepticism, check news sources, and be aware that what you see may not be real. As technology advances, our capacity for critical thinking must keep pace. The battle for truth in the digital age has just begun.
