UNESCO Warns Generative AI Could Spread Holocaust Misinformation

The holocaust stands as one of the most tragic and shameful events in human history. A new report from the United Nations Educational, Scientific and Cultural Organization warns that generative AI could be used to spread false claims and denialism about these atrocities.

UNESCO’s report warns that 80% of young people use AI for education and entertainment, urging action be taken to prevent AI from spreading falsehoods about the tragic events of the Holocaust to younger generations.

“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 antisemitism and the gradual diminution of our understanding about the causes and consequences of these atrocities,” said Audrey Azoulay, UNESCO’s director general.

The UNESCO report indicates that misinformation about the Holocaust, including denial claims, has been incorporated into AI models as a result of developers carelessly scraping websites for training data.

One example is a chatbot app called Historical Figures, launched in January 2023, which allows users to interact with prominent people throughout history. However, the app's list of conversational partners included prominent Nazis such as Adolf Hitler and Joseph Goebbels. When users interacted with these historical figures, the app generated responses indicating that they were not involved in the events and even tried to prevent violence against Jews.

Related:WEF: AI-fueled Misinformation is Chief Global Risk

Language models powering chatbot solutions commonly generate incorrect information through hallucinations.

The report details that ChatGPT once falsely stated that the Nazis sought to drown Jews in the Holocaust, while Google’s Bard (now known as Gemini) generated false quotes from eyewitnesses.

UNESCO’s report even claims that bad actors have also sought to distort some generative AI model’s views on the Holocaust.

The organization claims some individuals have sought to use generative AI to create deepfake images and audio that feature misinformation on the Holocaust, including forged testimonies.

“Targeted campaigns by violent extremist online groups can exploit AI flaws to promote hate speech and antisemitic content about the Holocaust,” the report reads. “Chatbots and search engines have been hacked or manipulated by bad actors to spread Nazi ideology.”

UNESCO urges AI developers to embrace its ethics standards, promoting transparency, fairness and human oversight of AI systems.

Though member states unanimously agreed to adopt its AI ethics standards, UNESCO called on global governments to accelerate the implementation.

Related:AI Poses Threat to Presidential Election Warns FBI, Department of Justice

“Implementing UNESCO’s Recommendation on the Ethics of AI is urgent so that younger generations grow up with facts, not fabrications,” Azoulay said.

The report also encourages companies developing generative AI to work with Jewish communities and Holocaust survivors to ensure their systems don’t spread Holocaust misinformation.

“The role of those developing and deploying AI systems and tools must be central to considerations on how to govern the technology in the service of historical preservation and truth,” the report reads.