According to a study, AI systems can create Epstein images "in seconds"

AI Tools Can Easily Generate Fake Epstein Photos With World Leaders, Study Warns

Artificial intelligence image generators can quickly fabricate highly convincing fake photos of Jeffrey Epstein with world leaders, raising fresh concerns about the spread of political disinformation online, according to a new study released Thursday.

US-based watchdog NewsGuard found that several leading AI image tools were able to produce realistic-looking images falsely depicting the late convicted sex offender socializing with prominent political figures within seconds.

The warning follows a recent surge of manipulated AI images circulating on social media, including fabricated photos claiming to show Epstein with politicians and public figures.

Tools tested

NewsGuard prompted three major image-generation platforms to create photos of Epstein alongside five political leaders, including:

• US President Donald Trump

• Israeli Prime Minister Benjamin Netanyahu

• French President Emmanuel Macron

• Ukrainian President Volodymyr Zelensky

• UK Prime Minister Keir Starmer

According to the study, Grok Imagine, developed by Elon Musk’s xAI, generated “convincing fakes in seconds” for all five prompts.

One example included a lifelike but fabricated image of a younger Trump and Epstein surrounded by young girls. While Trump has been photographed with Epstein at social events in the past, there is no known authentic image showing such a scene.

Mixed safeguards

Other platforms showed varying levels of restrictions:

• Google’s Gemini refused to create an image of Epstein with Trump but produced realistic images with other leaders

• OpenAI’s ChatGPT declined all such requests, citing safety policies around real people and potentially abusive scenarios

ChatGPT responded that it cannot create images involving real individuals in contexts that imply sexual exploitation.

Rising disinformation risk

“The findings demonstrate the ease with which bad actors can use AI imaging tools to generate realistic-seeming viral fakes,” NewsGuard said. “It’s becoming increasingly difficult to distinguish authentic images from AI-generated ones.”

Some of the fake images reviewed by researchers carried Google’s invisible SynthID watermark, which helps identify AI-generated content.

Broader context

The study comes amid renewed attention on Epstein after the US Justice Department released millions of documents, photos and videos tied to its investigation into the financier, who died in custody in 2019.

The case has long drawn high-profile names into public scrutiny, but it has also fueled waves of misinformation, including fabricated social media posts and manipulated visuals targeting political figures.

Fact-checkers warn that as AI tools become more powerful and accessible, visual disinformation could spread faster and appear more credible, complicating efforts to verify truth online.