Deepfake Celebrity Investment Scams: How Fake Endorsement Videos Work

A short clip runs as a paid social ad

A video ad, often just 15–30 seconds, shows a recognizable face making an urgent pitch — placed through hijacked ad accounts or newly created pages designed to look like a news outlet.

The face is real, the words never were

Face-swap and lip-sync AI models are layered onto real footage of a public figure, generating new audio and mouth movement to say things that person never said.

A fake news segment adds false legitimacy

Clips are often edited to look like they aired on a real broadcast — network logos, a studio backdrop, a chyron — to borrow credibility from a news brand the viewer already trusts.

The ad links to a slick, professional-looking platform

Landing pages mimic real brokerages or exchanges, complete with fake regulatory badges, fabricated user testimonials, and a live-looking "portfolio" that only ever goes up.

A small first deposit is designed to "work"

An initial deposit shows fast, real-looking gains in the dashboard — building trust before the platform blocks withdrawals or demands a "tax" or "unlock fee" to release funds.

A fake support agent keeps the story going

Once a victim tries to withdraw, a "customer success" contact appears with a plausible-sounding reason the money is temporarily held, stalling until further deposits are made.

The same footage runs across thousands of near-identical ads

A single deepfake clip is repurposed across dozens of ad accounts and fake pages simultaneously, so taking one down rarely stops the campaign.

The endgame is the deposit, not the platform

There is often no real trading happening at all — the dashboard numbers are generated locally in the browser, and every dollar deposited goes straight to the operator.

Public interviews and conference talks, re-voiced

Source footage is usually a real, publicly available interview or keynote — the original words are stripped out and replaced with an AI-cloned voice reading the scam script.

Real news broadcasts, edited and captioned

Genuine segments from major outlets are cut, re-captioned, and paired with a fabricated headline implying the celebrity or network endorsed a specific platform.

Fully synthetic 'interview' footage

Some campaigns generate an entire fake interview scene from scratch — a synthetic anchor, a synthetic guest, and a set that never existed, built end-to-end with AI video tools.

Stolen brand assets for the landing page

Logos, color schemes, and layouts are copied from real, regulated brokerages to make the destination site look established rather than freshly registered.

Fabricated regulator badges and licenses

Platform pages display invented or copied financial-authority seals that link nowhere, or link to a real regulator's homepage rather than an actual license lookup.

Recycled testimonials with stock or stolen photos

Customer reviews on the platform use stock photography or images lifted from unrelated real people's social profiles, paired with invented profit numbers.

Cloned voices from short public audio clips

A few seconds of a public figure's voice from any published video is often enough to train a convincing voice clone for the fake narration.

Hijacked verified accounts to distribute the ad

Compromised or purchased social accounts with existing follower counts and verification badges are used to post or boost the video, adding a false signal of trust.

Guaranteed or unrealistic returns

Any claim of guaranteed daily profit, "AI trading bot" with no losing days, or doubling your money in weeks is a hard stop regardless of who appears to be saying it.

Mismatched lip movement or unnatural blinking

Face-swap video often shows a slight lag or blur around the mouth, an unnatural jaw movement, or blinking that doesn't quite match a real, unscripted interview.

A voice that's almost right but flattened

Cloned audio frequently loses the natural pauses, breathing, and pitch variation of real speech, sounding slightly monotone or oddly paced compared to known footage of that person.

The platform name doesn't appear anywhere in real coverage

Searching the exact platform name alongside the celebrity's real name turns up no legitimate news results — only the ad itself and copies of it.

Urgency and scarcity language

Phrases like "limited spots today," "offer ends at midnight," or "before the government shuts this down" are designed to prevent you from pausing to verify.

The ad links off-platform immediately

Clicking through leaves the social platform entirely for a page with no working support contact beyond a chat widget or a single messaging-app handle.

No verifiable regulatory registration

The platform is not listed on your country's actual financial regulator's public register, even though it displays a badge claiming to be licensed.

The account posting the ad has no real history

The page or account behind the ad was created recently, has little unrelated content, or was clearly repurposed from an unrelated brand — a common sign of a hijacked account.

Check the celebrity's own verified channels

Public figures targeted by these campaigns frequently post direct warnings on their own verified accounts — search their name plus the platform name to see if they've already denounced it.

Reverse-search any face used in testimonials or the landing page

Run profile photos or "customer" images through a reverse face search — stock photography or a stolen photo tied to an unrelated real person confirms the testimonials are fabricated.

Look up the platform on your country's actual regulator register

Verify any claimed license number directly on your national financial regulator's public lookup tool, not the badge displayed on the platform's own site.

Search the video's exact wording

Paste a distinctive line from the pitch into a search engine — the same script recycled across dozens of "different" celebrity ads is a clear tell.

Try a real deepfake detection pass

Slow the video down and watch the mouth, eyes, and edges of the face closely for blur, flicker, or misalignment with the background — most current fakes still show seams on close inspection.

Never deposit through a link from an unsolicited ad

Go directly to a brokerage or exchange you already know and trust rather than following a link from a social media ad, no matter how convincing the endorsement looks.

Start with nothing you can't afford to lose entirely

Treat any first deposit on an unfamiliar platform as money you're comfortable losing completely before trusting it with more.

Ask in an unrelated, established investing community

Post the platform name in a well-known finance or investing forum and ask if anyone recognizes it — these campaigns get flagged quickly once enough people compare notes.

Stop sending any further funds immediately

Every additional request after a deposit — a "tax," a "verification fee," or an "unlock fee" to withdraw — is a continuation of the same scam, not a real obstacle.

Contact your bank or payment provider right away

Ask about disputing or reversing a card payment, wire transfer, or crypto transaction — acting quickly materially improves recovery odds.

Document everything before the site disappears

Screenshot the ad, the landing page, deposit confirmations, and any chat with "support" before the platform is taken down or you're blocked.

Report the ad to the platform it ran on

Social platforms have dedicated reporting flows for scam and impersonation ads — flagging it can get the ad account suspended and reduce further reach.

File with your national financial regulator and fraud agencies

In the US, report to the SEC, CFTC, and FTC/IC3; other countries have equivalent financial-conduct regulators and cybercrime units worth notifying directly.

Be wary of "recovery" services that appear afterward

A second wave of scammers specifically targets people who lost money in the first scam, offering fake fund-recovery services for an upfront fee.

Report to IC3 if crypto or cross-border transfers were involved

The FBI's Internet Crime Complaint Center tracks these operations and can link isolated reports to larger, known networks running the same footage.

Warn your network before they see the same ad

Post a factual warning wherever you saw the ad — these campaigns run at scale, and a public record helps the next targeted person recognize it faster.