It took just seconds to scroll past the image — two men in a pool, one of them Canada’s new prime minister, the other the billionaire financier whose name has become shorthand for depravity. But that image of Mark Carney with Jeffrey Epstein wasn’t real, and it wasn’t innocent. It was an AI-generated fabrication designed to smear, and its unmasking offers a rare look at how academic detection tools are being deployed against deepfake political attacks.

4 related posts

Debunking organization: Lead Stories with UB Media Forensic Lab · Detection tool: DeepFake-o-Meter · Date of first fact-check: January 28, 2025 · Subject of debunked image: Mark Carney and Jeffrey Epstein in swimming pool · Verdict: AI-generated / fake

Quick snapshot

1Confirmed facts
2What’s unclear
3Timeline signal
  • Image first circulated early 2025 (AI Incident Database)
  • Lead Stories fact-check published January 28, 2025 (University at Buffalo School of Engineering)
  • Conservative Instagram account posted it March 10, 2025 (NewsGuard Reality Check)
4What’s next
  • Origin of deepfake still being investigated (AI Incident Database)
  • Broader pattern of deepfake attacks on Canadian politicians may accelerate (NewsGuard Reality Check)
  • Detection tools like DeepFake-o-Meter likely to play larger role in elections (University at Buffalo School of Engineering)

Here is a quick reference of the key facts.

Fact Details
Date of first fact-check January 28, 2025
Debunking organization Lead Stories
Detection tool DeepFake-o-Meter (University at Buffalo)
Additional fact-checks Reuters (March 26, 2025), Yahoo News (February 26, 2026), The Walrus (April 10, 2026)
Subject of debunked image Mark Carney and Jeffrey Epstein in a swimming pool
Verdict AI-generated / false

What is the latest verified information about Mark Carney and Jeffrey Epstein?

Deepfake image debunked by Lead Stories

On January 28, 2025, the fact-checking organization Lead Stories (a fact-checking organization) published an analysis concluding that the image of Mark Carney with Jeffrey Epstein in a swimming pool is not authentic and is likely AI-generated. The claim that Carney had any association with Epstein is false.

Use of UB’s DeepFake-o-Meter in analysis

Lead Stories used the DeepFake-o-Meter tool developed by the University at Buffalo’s Media Forensic Lab (academic research lab) to assess the image. The tool detected manipulation artifacts consistent with AI generation.

No credible evidence of any connection between Carney and Epstein

Multiple fact-checking outlets, including Reuters (global news agency) and Yahoo News (news aggregation site), have confirmed the image is fake. There is no verified evidence of any relationship between Carney and Epstein. The closest apparent link is a 2013 photo of Carney standing next to Ghislaine Maxwell, as reported by NewsGuard Reality Check (misinformation monitoring service).

Why this matters

A fake image that implies a Canadian leader was involved with Epstein’s sex trafficking network is not just a lie — it’s a weapon. For voters and journalists, the pattern is clear: deepfakes are no longer hypothetical.

The implication: even high-quality debunking may not catch up to the speed of viral spread. The image first appeared early 2025 and was shared again after Carney became Liberal Party leader, according to AI Incident Database (a repository tracking real-world AI harms).

What should readers know first about the Mark Carney Jeffrey Epstein link?

The viral image is entirely fabricated

The image bearing signs of digital manipulation, including a woman’s wrist that was improbably curved and Carney’s hand appearing misshaped, according to NewsGuard Reality Check (misinformation monitoring service). These are classic tells of AI-generated content, as explained by Full Fact (UK-based independent fact-checking organization).

No legitimate source supports a relationship between the two

Neither court records, flight logs, nor any credible journalistic investigation has ever linked Carney to Epstein. The false image was promoted by Canadian and U.S. conservative accounts, according to the AI Incident Database (a repository tracking real-world AI harms).

Political motivations behind the deepfake are suspected

NewsGuard Reality Check (misinformation monitoring service) reported that a conservative Instagram account posted the image on March 10, 2025, claiming Carney was the first Canadian prime minister with known Epstein connections — a claim with zero factual basis.

The catch

Even after debunking, the image continues to circulate in private messaging groups and on fringe platforms. For readers, the best defense is skepticism: if an image looks too damaging to be true, it probably isn’t real.

Which official sources confirm key claims about Mark Carney and Jeffrey Epstein?

Three authoritative sources underpin the debunking of this deepfake.

Source Role Key finding
University at Buffalo School of Engineering Academic research lab Lead Stories used DeepFake-o-Meter; image not authentic
NewsGuard Reality Check Misinformation monitoring service Image showed signs of manipulation; posted by conservative account
Reuters Global news agency Debunked a related AI image of Carney, Maxwell, and Tom Hanks

Additionally, Full Fact (UK-based independent fact-checking organization) notes that AI tools alone are not definitive for verification, so reverse image search and plausibility checks remain gold standards. The pattern: academic and journalistic collaboration worked effectively here, but it’s reactive, not proactive.

What is still unclear or unverified about the Mark Carney Jeffrey Epstein connection?

Origin of the deepfake image remains unknown

According to AI Incident Database (a repository tracking real-world AI harms), the identity of the person or group who created the deepfake is not publicly known. The database frames the episode as harming epistemic integrity and the Liberal Party.

Motivation and creator behind the smear campaign

It is unclear whether the image was part of a coordinated disinformation campaign, though the timing — coinciding with Trudeau’s resignation and Carney’s leadership win — suggests strategic intent, as noted by NewsGuard Reality Check (misinformation monitoring service).

Potential role of automated bot networks in spreading the image

No official statement from Mark Carney’s office has been released regarding the deepfake, leaving room for continued uncertainty. The AI Incident Database (a repository tracking real-world AI harms) notes that the false image harmed not only Carney but the Canadian public’s trust.

The question: as detection improves, will attribution catch up? For now, the originator of the smear remains faceless.

What are the most common user questions on Mark Carney and Jeffrey Epstein?

Did Mark Carney actually meet Jeffrey Epstein?

There is no verified evidence of any meeting. The 2013 photo of Carney with Ghislaine Maxwell — republished by the Toronto Sun in January 2025, according to NewsGuard Reality Check (misinformation monitoring service) — is the closest link, but Maxwell and Epstein are not the same person.

Is the photo of them together real?

No. Multiple fact-checkers, including Reuters (global news agency) and Yahoo News (news aggregation site), have confirmed the image is AI-generated. Full Fact (UK-based independent fact-checking organization) suggests looking for errors in fingers, ears, and teeth — all present in this image.

How can I tell if an image is a deepfake?

Full Fact (UK-based independent fact-checking organization) recommends checking for unnatural limbs, inconsistent lighting, and missing watermarks (though some AI tools remove them). Tools like DeepFake-o-Meter can provide additional analysis, but human judgment remains critical.

The trade-off

Detection tools are powerful but imperfect. For a reader, the simplest test is context: if a photo surfaces that aligns perfectly with a political narrative — especially one that smears a candidate — treat it as suspect until verified by at least two independent sources.

Timeline of the deepfake’s spread and debunking

  • Early 2025: AI-generated image of Mark Carney with Jeffrey Epstein begins circulating online (AI Incident Database).
  • January 28, 2025: Lead Stories publishes fact-check using UB’s DeepFake-o-Meter, concluding the image is fake (University at Buffalo School of Engineering).
  • March 10, 2025: Conservative Instagram account posts the image (NewsGuard Reality Check).
  • March 26, 2025: Reuters debunks another AI-generated image showing Carney, Ghislaine Maxwell, and Tom Hanks on a beach (Reuters).
  • February 26, 2026: Yahoo News reports on the fake image and its debunking (Yahoo News).
  • April 10, 2026: The Walrus publishes an article discussing Canada’s links to Epstein, referencing the Carney deepfake (The Walrus (Canadian magazine)).
Bottom line: The Carney-Epstein image was debunked within weeks, but its spread shows how easily AI-generated smears can outpace fact-checks. For Canadian voters: demand verification before sharing. For platform companies: automated detection must become standard before the next election cycle.

Confirmed facts

  • The image of Mark Carney with Jeffrey Epstein in a swimming pool is AI-generated (University at Buffalo School of Engineering).
  • Lead Stories used UB’s DeepFake-o-Meter to detect the manipulation (University at Buffalo School of Engineering).
  • Multiple credible news outlets (Reuters, Yahoo) have debunked the image (Reuters; Yahoo News).
  • There is no verified evidence of any relationship between Carney and Epstein (NewsGuard Reality Check).

What’s unclear

  • Who created the deepfake image (AI Incident Database).
  • The specific motivation or target of the disinformation (AI Incident Database).
  • Whether the image was part of a larger coordinated campaign (NewsGuard Reality Check).
  • If Mark Carney’s office will issue a formal statement (NewsGuard Reality Check).

“The DeepFake-o-Meter looks for subtle inconsistencies in pixel patterns that are invisible to the human eye. In this case, it flagged artifacts consistent with AI generation.”

— Siwei Lyu, director of the Media Forensic Lab at the University at Buffalo, via University at Buffalo School of Engineering

“The image is not authentic and is likely an AI creation. We found no evidence that the depicted event ever occurred.”

— Lead Stories fact-check editors, as summarized by University at Buffalo School of Engineering

For Canadian voters, the decision is clear: dismiss any unverified image of a political figure with Epstein, and rely on established fact-checkers. For social media platforms, the implication is equally direct: invest in proactive detection, or face a crisis of trust far deeper than any single deepfake.

For readers unfamiliar with the background, Jeffrey Epsteins biography and scandal provides essential context on the financier’s life and the scandal surrounding him.

Don't miss these

Frequently asked questions

What tools are used to detect deepfakes?

Tools like the University at Buffalo’s DeepFake-o-Meter analyze pixel-level artifacts. Other tools include reverse image search and watermark detection (e.g., SynthID).

What should I do if I see a suspicious image?

Do not share it. Use reverse image search (Google Images, TinEye). Check fact-checking sites like Lead Stories, Reuters, or Full Fact. Report the post on the platform.

Why do deepfakes go viral so quickly?

They tap into existing biases and emotional triggers. The Carney-Epstein image exploited public anger toward elites, making it highly shareable before verification could catch up (AI Incident Database).

Has the creator of the Carney-Epstein deepfake been identified?

No. The identity remains unknown, and it is unclear whether the image was created by an individual or a coordinated group (NewsGuard Reality Check).

What are the legal consequences for creating such deepfakes?

In Canada, deepfakes used to harm reputation or interfere with elections could violate defamation laws or the Canada Elections Act. However, enforcement is challenging when creators are anonymous and based abroad.

As deepfakes become more sophisticated, the responsibility falls on both voters and platforms to remain vigilant against AI-powered disinformation.