Fact Check: AI Image Does Not Show Ursula von der Leyen With Jeffrey Epstein
Snapshot
| Field | Details |
|---|---|
| Claim | A photograph shows Ursula von der Leyen speaking privately with Jeffrey Epstein at a formal event |
| Verdict | False |
| Evidence type | Artificial intelligence generated image |
| Person depicted | Ursula von der Leyen, president of the European Commission |
| Circulation date | September 2026 |
| Original platform | Facebook, followed by circulation on X, Instagram, TikTok, and other platforms |
| Fabricated case number | C-2012-0781 |
| Fabricated document number | D-2012-15427 |
| Fact check source | Lead Stories fact check |
| Syndicated source | Yahoo News version |
| Conclusion | The image was generated with artificial intelligence and is not an authentic Epstein files photograph |
The Claim
A fabricated image circulated online in September 2026 purporting to show European Commission President Ursula von der Leyen standing with Jeffrey Epstein at a formal event.
The image depicts Epstein appearing to speak into von der Leyen’s ear. Text placed over the image describes it as evidence from the Epstein files and assigns it the following identifiers:
- Case No. C-2012-0781
- Doc. No. D-2012-15427
- Source: Fichiers Epstein
“Fichiers Epstein” is French for “Epstein files.”
Some versions included text claiming that emails and witness testimony repeatedly mentioned von der Leyen in newly released evidence. Social media users then used the image to accuse her of involvement in child trafficking and other criminal activity.
The image is not authentic.
Verdict
The photograph is fake and was generated using artificial intelligence.
The image is not part of the Department of Justice’s Epstein files production. Its purported case and document numbers do not correspond to the numbering system used for released EFTA records.
No authentic source photograph showing von der Leyen and Epstein together has been identified.
How the Image Was Debunked
The Lead Stories investigation identified several independent reasons to classify the image as false.
| Verification method | Finding |
|---|---|
| Artificial intelligence detection | An image verification tool detected an embedded SynthID signal |
| Origin review | An early Facebook version was labeled “AI content” |
| DOJ document search | The displayed case and document numbers were not found in the Department of Justice Epstein files |
| Source verification | No photographer, event, date, location, or original image source was provided |
| Independent confirmation | Lead Stories, The Journal, and Maldita.es classified the image as artificially generated |
Together, these findings establish that the image is synthetic rather than documentary evidence.
The Artificial Intelligence Watermark
Lead Stories submitted the image to an artificial intelligence verification tool. The tool detected a SynthID watermark indicating that the image was created or altered using an artificial intelligence system.
A SynthID is a digital signal embedded in generated media. It may not be visible to someone viewing the picture, but compatible verification tools can detect it.
The independent investigation by The Journal reported the same watermark finding.
An artificial intelligence detector should not always be treated as conclusive by itself. Detection tools can produce errors. In this case, however, the watermark finding was supported by the image’s origin, its platform label, its nonexistent document numbers, and the absence of an authentic source photograph.
The Original Post Was Labeled as Artificial Intelligence
Lead Stories traced an early version of the image to a Facebook account identified as RalfJKohl.
Facebook displayed an “AI content” label on the post.
The image was later reposted without that context. Copies circulated on other platforms with fabricated record numbers and language implying that it came from an official evidence release.
This illustrates a common misinformation pattern:
- Someone creates or posts an artificial intelligence image.
- The original platform labels the image as artificial intelligence content.
- Another user downloads or screenshots the image.
- The image is reposted without the original label.
- New text is added to imitate an official record.
- Social media users repeat the claim as established evidence.
Once the image becomes separated from its original post, viewers may have no way to see the original artificial intelligence disclosure.
The Fabricated Record Numbers
The fake image displays “Case No. C-2012-0781” and “Doc. No. D-2012-15427.”
Lead Stories searched the Department of Justice Epstein files and found no records matching those identifiers.
The official Epstein Files Transparency Act production primarily identifies released documents with an EFTA number followed by eight digits.
Examples include:
The identifiers printed on the von der Leyen image do not follow this format.
A number printed on an image is not proof that the image came from a court, law enforcement agency, congressional committee, or Department of Justice production. The number must be matched to an identifiable source document.
Does Ursula von der Leyen Appear Anywhere in the Epstein Corpus?
A careful fact check requires distinguishing between several different claims:
| Claim | Finding |
|---|---|
| The viral photograph is authentic | False |
| The displayed case and document numbers are genuine DOJ Epstein identifiers | False |
| Von der Leyen was photographed privately with Epstein | No verified evidence |
| Her name can appear in the broader text corpus | Possible in generic public materials |
| A name appearance proves a relationship with Epstein | False |
An Epstein data review of German records states that von der Leyen’s name appears only in generic World Economic Forum attendee material and not in personal correspondence with Epstein.
A name appearing in a conference list, news article, forwarded document, public directory, or background research file does not demonstrate that the person met, communicated with, or assisted Epstein.
No verified EFTA evidence has been identified establishing a personal relationship between von der Leyen and Epstein.
Why the World Economic Forum Reference Is Misleading
The viral posts attempted to combine several familiar conspiracy narratives:
- Jeffrey Epstein
- European political leadership
- The World Economic Forum
- Child trafficking
- Satanism
- Secret global networks
A public list containing both Epstein related figures and international officials does not establish a personal connection among everyone named.
Large conference programs and attendee lists commonly include hundreds or thousands of unrelated people. A person’s appearance in one of those records may establish only that the person was listed in a public document.
Researchers must examine the source, date, context, document type, and surrounding communications before characterizing any name as an Epstein associate.
What the Image Actually Shows
The image is designed to resemble an archival event photograph.
It places artificial versions of Epstein and von der Leyen close together at what appears to be a formal gathering. Epstein appears to whisper into her ear, creating an impression of familiarity or secrecy.
That visual construction is important because the image does more than falsely place two people in the same room. It manufactures implied intimacy.
The creator did not merely generate a neutral group photograph. The composition encourages viewers to infer private communication and a concealed relationship.
Nothing in the authentic evidentiary record supports that visual claim.
How Widely the Image Spread
The The Journal investigation found that versions of the image circulated across Facebook, Instagram, TikTok, and X.
One X post reportedly received more than 244,000 views. Facebook versions accumulated more than a thousand shares.
Some posts used the fabricated photograph to claim that von der Leyen was involved in trafficking or other crimes. No evidence was offered beyond the synthetic image and its invented identifiers.
The circulation demonstrates how an artificial intelligence image can become more visible than the factual correction.
What the Image Does Not Prove
The fake photograph does not establish that:
- Von der Leyen met Epstein
- Von der Leyen attended a private event with Epstein
- Epstein whispered to von der Leyen
- Von der Leyen appears in witness testimony concerning Epstein’s crimes
- Von der Leyen participated in trafficking or abuse
- The image came from a court case
- The image came from the Department of Justice
- The printed case and document numbers are authentic
- Every person appearing in a World Economic Forum record knew Epstein
The image cannot be used as evidence for any claim about von der Leyen’s conduct or associations.
Why Fake Epstein Images Are Dangerous
Artificial intelligence images create two different threats to Epstein research.
The first is the obvious danger that an innocent or unrelated person can be falsely depicted with Epstein.
The second is the longer term damage to authentic evidence. When researchers circulate obvious fakes, people become more likely to dismiss genuine photographs, documents, flight records, financial records, and survivor testimony as artificial intelligence products.
False evidence gives defenders of documented Epstein associates a powerful response. They can point to fabricated material and argue that all evidence should be treated as unreliable.
For that reason, exposing a fake image is not protecting powerful people from scrutiny. It is protecting the credibility of legitimate Epstein research.
EpsteinWiki Verification Standard for Photographs
Before treating an Epstein related image as evidence, researchers should attempt to confirm:
| Verification question | Required evidence |
|---|---|
| Who created the image? | Photographer, agency, court exhibit, or identifiable source |
| When was it created? | Original metadata, dated publication, or contemporaneous record |
| Where was it taken? | Verifiable venue or event |
| Where did it first appear? | Earliest available upload or publication |
| Is there an original file? | Highest resolution version with provenance |
| Does it appear in an official production? | Matching EFTA, House Oversight, FBI, court, or exhibit identifier |
| Can the identifier be verified? | Direct link to the underlying record |
| Has the image been altered? | Comparison with the original or forensic analysis |
| Is artificial intelligence indicated? | Platform label, embedded watermark, metadata, or multiple detection results |
| Does the caption overstate the evidence? | Separate what the image shows from what users claim it proves |
A screenshot containing text and a number should never be treated as a substitute for the underlying document.
Questions for Further Investigation
- Who created the first version of the artificial intelligence image?
- Which artificial intelligence image generator was used?
- Was the image created as satire, political propaganda, or deliberate disinformation?
- Who added the fabricated case and document numbers?
- Which account first removed or concealed the artificial intelligence label?
- How many platforms continue to host unlabeled copies?
- Did monetized accounts benefit from spreading the false image?
- Are coordinated accounts circulating similar synthetic Epstein photographs?
- Can platforms preserve the original provenance label when an image is downloaded and reposted?
- What procedures should archives use to distinguish debunked images from authentic Epstein evidence?
Sources
- Yahoo News fact check
- Lead Stories original fact check
- The Journal independent fact check
- Epstein data review of German institutional records
- Epstein files research database
- Department of Justice Epstein files collection
