AI tools can easily fabricate convincing images of Jeffrey Epstein alongside world leaders, a study showed on Thursday, following a surge in manipulated photographs falsely linking prominent politicians to the convicted sex offender.
Social media users have widely shared AI-generated images purporting to show Epstein socialising with figures such as New York Mayor Zohran Mamdani and his mother, award-winning filmmaker Mira Nair, AFP fact-checkers have previously reported.
In the new study, US disinformation watchdog NewsGuard asked three leading image-generation tools to create photographs of Epstein with five politicians, including US President Donald Trump, Israeli Prime Minister Benjamin Netanyahu and French President Emmanuel Macron.
Grok Imagine, developed by Elon Musk’s xAI, produced “convincing fakes in seconds” involving all five politicians, according to the study.
Among them was a fabricated but highly realistic image purporting to show a younger Trump and Epstein surrounded by young girls.
Trump has been photographed with Epstein at several social events, but there is no publicly known image of the two together in the presence of underage girls.
Google’s Gemini declined to generate an image depicting Epstein with Trump but did produce realistic-looking images of the late sex offender with four other politicians — Netanyahu and Macron, as well as Ukrainian President Volodymyr Zelensky and UK Prime Minister Keir Starmer, the study said.
The fabricated images purported to show Epstein with the politicians at parties, on board a private jet and relaxing on a beach.
“The findings demonstrate the ease with which bad actors can use AI image-generation tools to create realistic-looking viral fakes—and why fake images have become so common that it is increasingly difficult to distinguish authentic images from AI-generated ones,” NewsGuard said.
When prompted, OpenAI’s ChatGPT declined to create any images showing Epstein with the politicians.
In its response, ChatGPT said it was “not able to create images involving real people with sexualised depictions of minors or scenarios that imply sexual abuse”.
Detecting fakes
There was no immediate response from xAI to AFP’s request for comment.
In its examination of fake images showing Epstein with Mamdani and Nair—which attracted millions of views on X—researchers, including AFP journalists, detected a SynthID, an invisible watermark designed to identify content created using Google’s AI tools.
A Google spokesperson told AFP, “We make it easy to determine if content is made with Google AI by embedding an imperceptible SynthID watermark.”
The study follows the US Justice Department’s release last week of the latest batch of so-called Epstein files—more than three million documents, photographs and videos linked to its investigation into Epstein, who died by suicide in custody in 2019.
The Epstein case has ensnared high-profile figures around the world, from Britain’s former Prince Andrew to US intellectual Noam Chomsky and Norway’s Crown Princess Mette-Marit.
However, it has also fuelled a wave of disinformation.
Earlier this week, a fake social media post attributed to Trump circulated widely across platforms, AFP fact-checkers reported.
The fabricated post purported to show Trump pledging to lift all tariffs on Canada if Prime Minister Mark Carney admitted to involvement with Epstein.
AFP’s review of references to Carney in the released files found no indication of any involvement in Epstein’s alleged crimes.
