Holocaust and social networks: between traditional negationism and false images generated by artificial intelligence

Like every year, a ceremony takes place this Sunday, September 6 at the Grande Synagogue de la Victoire in Paris, in memory of the Jewish deportees and the French victims of the Shoah. While there are only a handful of survivors left, the question of memory is increasingly being asked, especially as false information and videos on the subject proliferate on social networks.

At 101 years old, Ginette Kolinka, one of the last survivors of Auschwitz-Birkenau, does not know TikTok, X, or even Instagram, she who "doesn't even have the internet". She therefore does not see hatred online, but she is not surprised that it exists: "The problem with anti-Semitism and Holocaust deniers is that they are always there and unfortunately they will always be there."

At the heart of many online videos is Vincent Reynouard, a former teacher, convicted multiple times for contesting crimes against humanity. He made it a specialty. This is why, in his training, Romain Blandre, history professor in Strasbourg and academic correspondent for the Shoah Memorial, uses it to alert colleagues and students. "His business is to release Holocaust denial videos very regularly on social networks which are deregulated and which are not censored, like Gab and Odyssee. I am working on his denial of the assassination of the 86 Jewish victims at the Struthof. He makes a video of around twenty minutes to explain to his spectators that the Struthof gas chamber is not one, and that it is a simple cold room", he explains.

There is outright negationism, but also the distortion of history with the help of artificial intelligence. Among the false images created with the help of AI, which are widely circulating on the networks, there is one in black and white, it shows a deportee playing the violin, shirtless, his body starving. Nothing here corresponds to reality, not even the decor. "The barbed wire is installed haphazardly. And when it is installed haphazardly, we know it's a fake. There's no need to look any further," says Tal Bruttmann, a specialist in the history of the Shoah and in particular, the deciphering of photos from the period. He says he is worried about progress in artificial intelligence, because for historians this involves much longer verification work.

“The danger, in the medium term, is that we will have the production of true-false images, that is to say false and credible images.”

For the moment, Tal Bruttmann notes that most content related to the Shoah, which is created with AI, distorts reality by playing on emotion. Their creators are mainly looking to make money.

Despite everything, the aim is also often ideological, according to Karel Fracapane, who is in charge of questions of memory and education at UNESCO. “As far as falsification is concerned, it is an approach which is more subtle, but which still consists of deliberately representing historical facts which are erroneous, so it is really disinformation. Anti-Semitism is not immediately obvious and social networks will be the main vectors,” he believes.

Karel Fracapane hopes for a little more controls from now on. In 2022, he conducted a study on the subject: “We realized that on unmoderated networks such as Telegram, around 50% of content relating to the Shoah was anti-Semitic, which is enormous. On this network for example, it is 80% of content in the German language”. Faced with all this content, Romain Blandre, from the Shoah Memorial, has his own educational method. According to him, it is necessary to confront students very early on with anti-Semitic comments or images to immunize them in some way, like a vaccine.