15 Famous Deepfake Examples and What Each One Teaches Us

The fastest way to understand deepfakes is to study real ones. Every famous case below fooled real people, and every one left clues. These deepfake examples span entertainment stunts, multimillion-dollar frauds, election interference, and viral hoaxes, and each entry ends with the detection angle: what gave it away, and what it teaches.

The fastest way to understand deepfakes is to study real ones. Every famous case below fooled real people, and every one left clues. These deepfake examples span entertainment stunts, multimillion-dollar frauds, election interference, and viral hoaxes, and each entry ends with the detection angle: what gave it away, and what it teaches.

Quick answer: Famous deepfake examples include the deeptomcruise TikTok videos, the fake Zelensky surrender broadcast, the AI Pope in a puffer jacket, the Biden New Hampshire robocall, and the $25.6 million Arup video-call fraud. Each shows a different technique: face swap, voice clone, full AI generation, or live impersonation.

If you are new to the technology itself, start with what is a deepfake, then come back for the case files.

How We Chose These Deepfake Examples

Three rules shaped this list. First, every case is documented: each entry carries a year and a named source, because deepfake stories mutate in retelling. Second, everything here is safe for work and safe for a classroom; we cover scams, politics, and hoaxes, not abuse content. Third, each example teaches something different, across four techniques (face swap, voice clone, full AI generation, live video impersonation) and four motives (demonstration, fraud, politics, chaos).

We analyze deepfakes for a living, so each case includes the detection read, not just the headline. Lift these for your presentation; that is what the page is for.

Entertainment and Demonstration Deepfake Examples

1. The deeptomcruise TikTok Videos (2021)

In February 2021, a TikTok account called @deeptomcruise posted clips of Tom Cruise doing magic tricks and golfing. Millions watched before learning the truth: Belgian visual effects artist Chris Umé had face-swapped Cruise onto actor Miles Fisher, as reported by The Verge in March 2021. It became the moment deepfakes went mainstream.

Detection angle: the videos were nearly flawless because they combined AI with a professional impersonator and days of manual cleanup per clip. The lesson: top-tier fakes will not show obvious glitches, so provenance checks and detection software matter more than squinting at pixels.

2. The Obama "PSA" by Jordan Peele (2018)

In April 2018, BuzzFeed and director Jordan Peele's Monkeypaw Productions released a video of Barack Obama saying things he never said, with Peele's voice driving a lip-synced Obama. It was published openly as a warning about synthetic media.

Detection angle: the mouth region carried subtle blurring and the cadence was Peele's, not Obama's. Early lip-sync fakes concentrate artifacts around the mouth and jawline. The lesson endures: the video told you it was fake, and people still found it unsettlingly convincing.

3. David Beckham's Nine-Language Malaria Campaign (2019)

In April 2019, the "Malaria Must Die" campaign released a video of David Beckham appealing for malaria eradication in nine languages, most of which he does not speak. Synthetic video technology mapped donated voices onto his face, as covered by the BBC.

Detection angle: this is a consented, disclosed synthetic video, which is exactly why it matters. Technique alone does not make content malicious; consent and disclosure separate a health campaign from a scam using identical tools.

4. The AI Pope in the Puffer Jacket (2023)

In March 2023, an image of Pope Francis in a white Balenciaga-style puffer coat went massively viral. It was not a video deepfake but a fully AI-generated image made with Midjourney by a Chicago creator, who described making it to BuzzFeed News.

Detection angle: the giveaways were in the details: a warped crucifix chain, a smeared hand gripping what may be a coffee cup, melted glasses edges. Generated images fail at fine structure. An AI image detector flags images like this from statistical fingerprints even when the eyes miss the clues.

Deepfake Fraud Examples

These cases link to documented losses; for the aggregate numbers, see our deepfake statistics roundup.

5. The $25.6 Million Arup Video-Call Fraud (2024)

In February 2024, a finance employee in the Hong Kong office of engineering firm Arup joined a video call with what looked like the company's CFO and colleagues. Every other participant was a deepfake. The employee made 15 transfers totaling about $25.6 million (HK$200 million), as reported by CNN and the South China Morning Post.

Detection angle: the fakes survived a live group call, but the request pattern was the tell: an unusual, urgent, secret transfer instruction that bypassed normal process. Out-of-band verification (calling the CFO back on a known number) would have killed the scam in one minute.

6. The Jennifer DeStefano Fake Kidnapping Voice Clone (2023)

In April 2023, Arizona mother Jennifer DeStefano answered a call and heard her daughter sobbing, followed by a man demanding ransom. Her daughter was safe; the voice was an AI clone. DeStefano recounted the call in testimony to the U.S. Senate Judiciary Committee in June 2023.

Detection angle: seconds of public audio can clone a voice. The countermeasure is procedural, not auditory: hang up, call the real person directly, and agree on a family code word before you ever need one.

7. The Hyderabad WhatsApp Voice Clone Scam (2023)

In 2023, a man in Hyderabad, India received a WhatsApp call from someone who sounded exactly like a relative in Canada claiming an urgent emergency, and transferred money before discovering the fraud, as reported by the Times of India.

Detection angle: voice clone scams scale globally because they exploit the same script everywhere: a loved one, a crisis, a transfer. Urgency plus a payment request is the signature, regardless of how convincing the voice sounds.

8. Celebrity Investment Scam Ads: Musk and Others (2022-ongoing)

Deepfaked Elon Musk has pitched fraudulent crypto platforms in fake ads and pop-up "live" streams for years. One Ontario woman lost about $1.7 million to a Musk deepfake investment scheme, per reporting carried by Yahoo Finance Canada, and CBS News reported that Musk-related deepfakes contribute to billions in U.S. fraud losses, citing Deloitte's estimate that AI-enabled fraud could reach $40 billion in the U.S. by 2027.

Detection angle: these fakes recycle real interview footage with cloned audio, so lip-sync drift and flat intonation are common tells. The bigger tell is the offer: no billionaire is doubling your crypto. Full breakdown in our celebrity deepfakes guide.

Political Deepfake Examples

9. The Fake Zelensky Surrender Video (2022)

In March 2022, weeks into Russia's invasion, hackers placed a deepfake of Ukrainian President Volodymyr Zelensky telling soldiers to lay down their arms on the hacked Ukraine 24 news website. The BBC and Reuters covered its rapid debunking, and Meta removed it platform-wide.

Detection angle: this was a sloppy fake: the head was oddly sized and lit differently from the body, and the accent was off. It failed within hours partly because Zelensky immediately posted a real rebuttal video. Speed of authentic counter-evidence is a defense in itself.

10. The Biden New Hampshire Robocall (2024)

Two days before the January 2024 New Hampshire primary, thousands of voters got a robocall in Joe Biden's cloned voice telling them to stay home and "save your vote." The FCC proposed a $6 million fine against political consultant Steve Kramer, and carrier Lingo Telecom settled for $1 million, as reported by NPR and PBS in May 2024.

Detection angle: the audio quality was mediocre, but phone audio hides artifacts, which makes voice the easiest deepfake channel. The case directly triggered the FCC's ban on AI-generated voices in robocalls.

11. The Slovakia Election Audio (2023)

Two days before Slovakia's September 2023 parliamentary election, audio spread on Facebook of candidate Michal Simecka and journalist Monika Todova supposedly discussing how to rig the vote. Fact-checkers found it AI-generated, but it landed inside the pre-election quiet period when rebuttals were restricted, as Bloomberg reported.

Detection angle: the timing was the weapon. Audio-only fakes are harder to debunk than video, and releasing one when media cannot respond maximizes damage. Treat conveniently timed leaked audio as suspect by default.

12. India's Campaign Deepfakes (2024)

India's 2024 general election normalized synthetic campaigning at scale. The DMK party released videos of its leader M. Karunanidhi, who died in 2018, praising current leadership, and AI translation spread candidates' speeches across languages, as reported by Al Jazeera in February 2024.

Detection angle: much of this was open, even disclosed, synthetic media. The hard question it teaches: when a real politician's words are AI-translated into a language they do not speak, where is the line? Detection tools flag the synthesis; humans must judge the intent.

Viral Hoaxes and Misattributions

13. The Pentagon "Explosion" Image (2023)

On May 22, 2023, an AI-generated image of black smoke beside the Pentagon spread through verified accounts on X, and U.S. markets dipped briefly before officials debunked it, according to the Associated Press. No explosion occurred.

Detection angle: the building facade blurred into the fence, a classic generation artifact, and crucially, no second photo ever appeared. A real event near a major landmark produces dozens of angles within minutes. One image, one angle, zero witnesses: assume synthetic.

14. Celebrity Hoax Compilations on TikTok and YouTube (ongoing)

Whole channels now exist around fake celebrity moments: invented arrests, fabricated interviews, AI voice "leaks." In April 2026, detection firm Copyleaks reported a surge of deepfaked celebrity ads on TikTok impersonating stars like Taylor Swift and Jennifer Aniston, as covered by Bitdefender.

Detection angle: hoax compilations launder fakes through reposting, and each re-upload re-compresses the file and degrades detectable signals. Always trace a viral clip to its earliest posting before judging it, and check the original file when you can.

15. The Marco Rubio Voice Impersonation Campaign (2025)

In July 2025, the U.S. State Department warned that an impostor used AI to mimic Secretary of State Marco Rubio's voice and writing style, contacting foreign ministers and U.S. officials through text and the Signal app, as reported by the Associated Press.

Detection angle: no video, no audience, just targeted messages to a handful of powerful people. Deepfakes are moving from broadcast hoaxes to precision social engineering. Verification protocols for high-trust contacts are no longer optional, even at the top of government.

What These Deepfake Examples Have in Common

Look across all 15 cases and the same patterns repeat:

• Urgency. Transfer now, vote differently tomorrow, your daughter is in danger. Nearly every malicious case compresses your decision time.
• Borrowed authority. A CFO, a president, a celebrity, a pope. Fakes hijack faces you already trust so you skip verification.
• Emotion before evidence. Fear, greed, outrage. The content is engineered so sharing feels more natural than checking.
• Channel choice. Scammers pick channels that hide artifacts: phone audio, compressed social video, group calls with small video windows.

That is the detection engineer's read: the technical fake gets you to look, but the social engineering gets you to act. Spot the pressure pattern and you catch fakes that fool your eyes.

How to Verify a Suspected Deepfake Yourself

1. Find the source. Trace the clip to its earliest posting. Real events leave a trail; fakes appear from nowhere.
2. Look for corroboration. Major news has multiple angles and outlets. One lone clip of a big claim is a red flag.
3. Check the details. Hands, jewelry, glasses, backgrounds, and lighting boundaries fail first in fakes.
4. Slow the audio down. Cloned voices often have flat emotional range and odd breathing patterns.
5. Run it through a detector. Software reads statistical fingerprints humans cannot see. Most of the cases above carry signals a modern detector flags.

FAQ

What is the most famous deepfake?

The deeptomcruise TikTok videos from 2021 are the most famous deepfake example: the moment the public learned a fake could pass casual viewing, covered worldwide after The Verge identified creator Chris Ume.

What was the first deepfake?

The term comes from a Reddit user called "deepfakes" who posted face-swapped videos in 2017. Academic face-manipulation research predates that, but 2017 is when the technology and the name escaped the lab.

Are deepfake examples illegal to watch or share?

Watching is not illegal. Sharing depends on content and intent: fraud, election interference, and intimate-image abuse carry real penalties. See are deepfakes illegal for the legal map.

How much money have deepfake scams stolen?

Documented single cases include Arup's $25.6 million (CNN, 2024) and an Ontario victim's $1.7 million Musk-scam loss. Deloitte estimates AI-enabled fraud could reach $40 billion in the U.S. by 2027. Our deepfake statistics page tracks the totals.

Can these deepfake examples be detected by software?

Most, yes. Detection models flag the statistical fingerprints of synthesis even in convincing fakes. The honest nuance: brand-new generation techniques can lag detector updates, which is why source verification belongs alongside any software check.

Conclusion: Learn From Deepfake Examples, Then Test Your Eye

These 15 deepfake examples tell one story in four genres: the technology keeps improving, but the playbook barely changes. Urgency, authority, and emotion did the real work in almost every case, from a fake CFO in Hong Kong to a cloned president in New Hampshire. Study the cases, internalize the patterns, and verify before you share or pay.

This page is updated quarterly as new documented cases emerge. For the full verification stack, start with our deepfake detection guide, then create a free account and check your first suspicious file today.

Related reading