The videos of dead children made with artificial intelligence that horrify families on TikTok

[ad_1]

Bereaved parents voiced disgust at a macabre TikTok trend. AI videos often look nothing like the real person TikTok Bereaved parents have expressed disgust at a macabre trend on TikTok where artificial intelligence is used to create videos that depict child victims of murder. The videos – taken without the consent of their families – show crudely animated versions of children, some of them babies, telling how they were killed. Since April, many of these computer-generated videos have been posted on TikTok, although the company announced a ban on underage AI versions in March. However, a BBC investigation has found evidence that indicates there may be confusion among TikTok employees over platform policy. A TikTok content moderator told the BBC that four months after the ban the company had still not informed him of the rule change. Some victim videos remained on the platform for several months, others received millions of views. Bereaved family members in several countries spoke about the AI ​​versions of their deceased loved ones that appeared online. Baroness Beeban Kidron, an independent member of the British Parliament’s House of Lords, says the videos can be particularly disturbing for bereaved families because they feature simulations of their deceased family members telling their own story. “If you’re someone who was attached to that child, especially a parent or a family member, it’s an assault, an emotional assault,” she said. ‘Aggression of the worst’ It wasn’t long before these videos started to go viral and relatives of some of the dead children began to speak out. One of the first was a Canadian named Amélie Lemieux. In July 2020, her daughters – Romy, 6, and Norah, 11 – were murdered by her ex-husband, who later committed suicide. In April of this year, Lemieux was horrified when she received a TikTok that featured an AI-powered version of Norah talking about the murders. “It’s the worst aggression you can imagine,” Lemieux said. “I couldn’t watch until the end, and neither could my brother. He was screaming at me to stop.” Amélie Lemieux was especially concerned that her nieces and nephews might randomly come across her dead cousin’s video while browsing TikTok. Although many people reported the video to the app, it remained online. Eventually, after Lemieux gave a series of interviews recounting how disturbed she was by the video, the anonymous account she posted it to was deleted. Amélie Lemieux couldn’t find out who created the video, but she is shocked by the actions. “It’s incredibly shocking for someone to use something so disturbing, something that will leave us incomplete for the rest of our lives,” she said. “All they want are views, likes, shares. And people who leave comments. Personally, I can’t forgive that.” New technologyThe creation of these videos was made possible by the increasing online accessibility of AI software. You can now easily create and animate a basic human figure and provide a script with the lines, which will be played back with a computer-generated voice. Some of the faces in victim videos do not look like the real person. But others are based on genuine photos. The people who make these videos tend to try to remain anonymous. But the BBC identified one creator as a student in London named Ritul. On his YouTube channel he has been promoting what he calls “storytelling” videos as a way to quickly attract followers on TikTok. “I gained thousands and thousands of followers by posting videos telling stories just using AI,” he said in a video. “I gained 47,000 followers, and that was just in less than three weeks.” The BBC has reached out to Ritul, but he declined to comment. His TikTok account has been deleted. James Bulger AI Videos Appeared in Multiple Languages ​​PA MEDIA TikTok videos need to grab attention within seconds or users will keep scrolling. That’s why creators like Ritul have focused on making videos about the most extreme or high profile cases. The stories of some children have been the subject of multiple videos in several languages, including Portuguese, Spanish, French, German, Italian and Polish. Over a two-month period, the BBC uncovered 170 different TikTok videos featuring AI versions of James Bulger, the two-year-old British boy who was murdered by two older children in Merseyside, England, in 1993. TikTokO response The phenomenon made headlines in Britain in July, when James’ mother, Denise Fergus, told a newspaper that videos of her son were “beyond sickening”. TikTok conducted a mass takedown of the videos and issued a statement that said, “There is no place on our platform for disturbing content of this nature. We continue to remove content of this nature as we find it.” The removed James videos included two in Vietnamese which, combined, had four million views. The BBC spoke to a TikTok content moderator in Vietnam, who agreed to be interviewed on condition of anonymity. He said that two weeks after the removal of the videos, the company had still not officially announced that this type of content was banned. He told the BBC that he believes the videos were released under TikTok’s rules, which allow survivors of abuse to share their stories. So TikTok employees are not clearly informed about what the company’s policy really is? TikTok didn’t specifically respond on that point when we reached out to the company for comment. But he made it clear that the moderator was wrong to suggest that these videos could be allowed on the platform as survivor experiences – because they are computer-generated images of dead children, rather than genuine survivor experiences. It’s still easy to find this type of video on the app. But TikTok says it has always been very open about the fact that it can’t monitor all available content. Additional reporting by Thuong Le of BBC News Vietnam

[ad_2]

Source link