The United Nations has warned that AI technology is being used to create false stories about World War II atrocities, including Holocaust denial, which could lead to a dangerous rise in anti-Semitism.
UNESCO, the UN’s education and culture body, has urged governments and tech companies to introduce ethical safeguards for AI technology and for schools to educate about the risks of AI-generated content.
“AI technology is being exploited to spread Nazi ideology and create false narratives about the Holocaust,” said UNESCO Director-General Audrey Azoulay.
“If we allow the horrific facts of the Holocaust to be diluted, distorted, or falsified through the irresponsible use of AI, we risk the explosive spread of anti-Semitism.”
She warned that it could also lead to a “gradual diminution of our understanding about the causes and consequences of these atrocities.”
AI tools like ChatGPT and DALLE-2 have become widely used since their introduction in 2022, for tasks ranging from translating texts to providing customer service. However, there are serious concerns about their potential misuse.
Google’s Gemini model caused controversy in February when it generated images of ethnically diverse Nazi soldiers. Additionally, ChatGPT invented the concept of “Holocaust by drowning,” and Google’s Bard chatbot fabricated witnesses to support untruths about Nazi massacres.
Karel Fracapane, a UNESCO expert on Holocaust education, explained that these distortions of Holocaust history show how AI can undermine our relationship with truth and erode democratic culture. He noted that the rise of far-right politicians in Western Europe is part of the same process as the spread of hate speech online.
Fracapane emphasized that while education about the Holocaust has improved, the spread of Holocaust denial and Nazi ideas both online and in the real world remains a significant issue. “What’s in this report is a manifestation of what is happening in society — it leads to very real political consequences,” he said.
Among the alarming examples highlighted in the UNESCO report were an AI-generated video of Joseph Goebbels falsely claiming he had tried to save Jewish lives and a deepfake of “Harry Potter” actress Emma Watson reading from Adolf Hitler’s “Mein Kampf.”
Despite these issues, UNESCO’s report also noted positive uses for AI, such as categorizing and sorting testimonies to find new patterns and insights and using AI-powered education tools to create immersive learning experiences for young people.
However, Fracapane concluded that with the current state of technology, he views AI more as a menace than an opportunity.