AI technology is helping to create false stories about World War II atrocities including Holocaust denial, risking an "explosive spread of anti-Semitism", the UN warned Tuesday.
The UN's education and culture body Unesco called for governments and tech companies to introduce ethical safeguards around AI technology, and for schools to spread the word about the risks of AI-generated content.
Unesco's report highlighted instances where hackers had rigged chatbots to spread Nazi ideology, and others where bots dreamt up their own stories around the Holocaust.
"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," said Audrey Azoulay, Unesco's director-general.
She said it could also lead to the "gradual diminution of our understanding about the causes and consequences of these atrocities".
Controversy
AI tools that generate content – such as ChatGPT and DALLE-2 – exploded onto the scene in 2022.
Companies and institutions have rushed to adopt them for anything from translating texts to providing customer service.
But policymakers, campaigners and law enforcement warned from the start that these tools could be used to extort money, ruin reputations, influence elections and distort history.
Several tools have already publicly failed at history.
Google's Gemini model caused global controversy in February when it generated images of ethnically diverse Nazi soldiers.
Unesco also pointed out that ChatGPT had wholly invented the concept of "Holocaust by drowning", while Google's Bard chatbot had fabricated witnesses to support untruths about Nazi massacres.
Unesco's policy paper, published jointly with the World Jewish Congress to mark the International Day for Countering Hate Speech, called for "urgent action" from governments, tech companies and educators.
Karel Fracapane, a Unesco expert on Holocaust education, told French news agency AFP that distortions of Holocaust history showed how AI could upend our relationship with truth and "lead to a deep erosion of democratic culture".
Real political consequences
Fracapane said the current rising popularity of far-right politicians in Western Europe was part of the same process as the spread of hate speech online.
Although education around the Holocaust had improved markedly in recent years, he said the "red line" of spreading Holocaust denial and Nazi ideas had been smashed both online and in the real world.
"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 examples highlighted in the Unesco report was an AI-generated Joseph Goebbels claiming he had tried to save Jewish lives, and a deepfake of "Harry Potter" actress Emma Watson reading from Adolf Hitler's Mein Kampf.
However, Unesco's report also flagged positive uses for AI.
Algorithms could categorise and sort testimonies, helping to find new patterns and fresh insights.
And AI-powered education tools could bring immersive experiences to young people, the Unesco report said.
But Fracapane said that, with the state of the technology right now, he was leaning harder towards the "AI as menace" side of the debate rather than "AI as opportunity".
(with AFP)
