At first glance, Amelia, with her purple bob and pixie-girl looks, seems unlikely to be adopted by the far right as an increasingly popular meme.
But in recent weeks, memes and AI-generated videos featuring the fictional British teenager have proliferated on social media, particularly on X. In it, Amelia parrots right-wing and often racist talking points, combining glorification of stereotypical British culture with anti-immigrant and Islamophobic tropes.
She drinks beers in pubs, reads Harry Potter, and travels back in time to fight in some of Britain’s most famous battles. But she has donned an ICE uniform to violently deport immigrants, and has embraced rhetoric so extreme that even British far-right activist Tommy Robinson posted a video of her. It’s an impossible life for a high school girl.
But Amelia has another characteristic that makes her “memetic.” That said, she was originally created for a computer game two years ago as part of the British government’s anti-extremism program.
The game, called Pathways: Navigating Gaming, the Internet & Extremism, was developed by Shout Out UK (SOUK), a not-for-profit organization aiming to improve the public’s understanding of politics, as part of a learning package funded by the UK Home Office.
It aims to educate young people about the dangers of online radicalization and asks them to navigate six different scenarios using multiple-choice options. Users play as the cartoon character “Charlie,” who enrolls in a new school and befriends “Amelia,” who shares anti-immigrant ideas and disinformation, and tries to recruit Charlie to anti-immigrant groups and protests.
The game is relatively simple and the logical leaps in each scenario have attracted attention online, but SOUK CEO Matteo Bergamini told CNN that the game is “not meant to be played alone.”
Rather, he explained, it was intended to be part of a “broader learning package that allows teachers to foster more nuanced discussions about what is healthy and safe behavior and what is potentially dangerous and illegal.”
Bergamini said Amelia’s appearance “wasn’t particularly important,” but experts say Amelia’s status as a white, purple-haired girl who espoused far-right ideas inadvertently created an avatar that could be exploited by the online right-wing.
Siddharth Venkataramakrishnan, an analyst at the Institute for Strategic Dialogue, said she has “a lot of checks” on the group, which meme-izes everything online with its particular sarcastic tone.
Her role in the game embodies a broader “stereotype” that many right-wingers have about the British government, which they perceive to be “anti-white” and a “nanny state,” she told CNN.
And importantly, she is a beautiful woman who thinks like them. Venkataramakrishnan added that similar accounts “condemn immigrants as sexual predators and sexual deviants, while at the same time it is surprising that so much of the editorial content is highly sexualized.”
Asked for comment, an Interior Ministry spokesperson told CNN that the Prevent strategy “has turned around 6,000 people away from violent ideologies, deterred terrorists and kept our country safe.” The local government where the game was created has not yet responded to CNN’s inquiry asking if the game is still in use.
The meme first started circulating on January 9th after the Telegraph, a right-wing British newspaper, published an article titled “Video game ‘Prevent’ treats all teenagers like far-right extremists”. Mr Bergamini called the headline “misleading”.
The following day, right-wing news outlet GB News picked up the story, incorrectly reporting that Game “warns children who question mass immigration that they will be treated like terrorists.”
Bergamini said this was “complete misinformation” and that the game only referred children to anti-extremism programs if they were involved in illegal activity, stressing that it was not his opinion on the issue. Still, because of that impetus, and the realization that the British government-funded game was cracking down on teenagers’ political opinions, the Amelia meme spread like wildfire.
“I think I’m in love with Amelia,” one X user posted along with a screenshot of the game, which received more than 5 million views. One X community group called “For the Based, the Phenomenon,” or Amelia, had more than 11,000 members as of Thursday. According to CoinGecko, Elon Musk retweeted an Amelia meme last week, and by the end of last week, two Amelia meme coins existed.
Memes act as a kind of coded language, imbued with different layers of meaning depending on how much context the viewer has, making them integral to any political discourse. Venkataramakrishnan then noted that there is a “certain degree of reasonable deniability” because statements that imply hate speech can be defended as “just a joke.”
Because AI allows people to churn out content almost instantly, these memes and images can spread rapidly, not only within the country where they originate (in Amelia’s case, the UK), but also internationally.
Callum Hood, research director at the UK- and US-based nonprofit Center for Countering Digital Hate, said flooding the internet with content like this that aligns with the far-right’s worldview “really helps” the far-right.
User-generated videos featuring Amelia quickly adopted references to other far-right memes beloved around the world. In one AI-generated video, she holds out two pills and urges “Charlie” to take the “purple pill.” It references far-right claims and the famous scene in “The Matrix” where Neo, played by Keanu Reeves, chooses between taking the blue pill that will keep him blissfully ignorant, or the red pill that will reveal a disgusting, enlightened reality.
In another AI-generated video posted on
CNN has reached out to X for comment.
Amelia is a particularly viral example, but using AI to generate content was already a popular pastime for Britain’s online right. This allows them to “manufacture support from British characters that seem relatable and believable,” or simply to generate an image “that you want to put in people’s minds,” Hood told CNN.
He cited the example of a Facebook page “filled with images of credible-looking British pensioners and veterans speaking out about their concerns about immigration,” run from Sri Lanka, according to the platform’s own transparency information.
And if tech companies don’t clearly label such content as AI, it will be especially difficult for people to distinguish between real and AI images, Hood added.
“Our research suggests that a significant number of people treat these images as trustworthy…Especially when there is an effort to make this image appear trustworthy, we find comments that seem to suggest that users are responding to it thinking it is a real person.”
