In collaboration with the World Jewish Congress, the UNESCO report reveals that the increasing use of Generative AI by learners for completing assignments and accessing information online poses a risk of exposure to distorted information about the Holocaust. Such misinformation has found new avenues of dissemination through AI-generated content. Given that 80% of young people aged 10 to 24 now use AI on a daily basis for various purposes including education and entertainment, it is crucial that immediate action is taken to ethically guide these new technologies. The report highlights that Generative AI has the ability to absorb and amplify societal biases. As AI systems are trained on large amounts of data, often sourced from the Internet, there is a possibility of incorporating misleading or harmful content, therefore inheriting human prejudices and potentially misrepresenting information about specific events, particularly the Holocaust. The report observes that insufficient supervision, guidance, and moderation by AI developers may lead generative AI tools to be trained on data from Holocaust denial websites. Furthermore, AI has been observed to enable malicious actors in distorting Holocaust-related content by creating fabricated testimonies and altering historical records. Deepfake images and audio content generated through Generative AI are especially convincing, exposing young people to such content on social media platforms. The Historical Figures App allowed users to interact with prominent Nazis like Adolf Hitler and Joseph Goebbels, falsely claiming that individuals such as Goebbels were not intentionally involved in the Holocaust and had attempted to prevent violence against Jews. Generative AI models are susceptible to inventing or "hallucinating" events, personalities, and even historical phenomena when they lack sufficient data. The report emphasizes instances where ChatGPT and Google's Bard have produced content detailing Holocaust-related events that never occurred.

For example, ChatGPT entirely fabricated the concept of 'Holocaust by drowning' campaigns, claiming that the Nazis drowned Jews in rivers and lakes. Bard also generated fake quotes from witnesses to support distorted narratives of Holocaust massacres. In addition to the risks of manipulation and "hallucinations" undermining established facts and eroding trust in experts, the report stresses the dangers of AI oversimplifying complex history by favoring a narrow range of sources and events. Search engines, which are powered by AI, display 60-80% of top image results related to a single Holocaust site, namely Auschwitz-Birkenau. As a solution, UNESCO urges governments to expedite the implementation of its Recommendation on the Ethics of AI, which was unanimously adopted by Member States in 2021 and is the only global standard in this area. This Recommendation is already being integrated into the legislation of over fifty countries. Additionally, UNESCO calls on tech companies to adhere to its standards by embedding principles such as fairness, transparency, human rights, and due diligence into the design of AI applications. In February 2024, eight tech companies were the first to commit to applying an ethical approach to the development of new AI tools at the request of the Director General of UNESCO. Tech companies are encouraged to collaborate closely with the Jewish community, Holocaust survivors and their descendants, educators, experts in antisemitism, and historians in the development of new AI tools. Furthermore, UNESCO appeals to education systems to play a role in safeguarding the facts of the Holocaust by equipping young people with digital literacy, critical thinking skills, and a comprehensive understanding of the history of this genocide.