Uruguay's Most Cloned Faces: 4,000 Deepfake Scam Ads Impersonating the Central Bank Chief, dLocal's Founder, and Lecueder

Between January and July 2026, Revelum detected nearly 4,000 fraudulent ads targeting Uruguay that used AI to clone the voices and images of the country’s best-known public figures. Uruguay’s El Observador published the findings on July 26, 2026, in a report by Juan Pablo De Marco.

The findings at a glance:

• ~4,000 fraudulent ads targeting Uruguay, with data running through July 22, 2026
• 441 ads cloned Guillermo Tolosa, president of the Banco Central del Uruguay — the most impersonated Uruguayan
• 374 ads cloned businessman Carlos Lecueder in eight days
• 62% of the ads sold fake investment platforms
• 33 → 1,244 monthly detected ads from January to June, nearly a 38x increase
• US$250, or 7,000–11,000 Uruguayan pesos, is the “activation” deposit the scripts request

---

Who Was Cloned: Uruguay’s Most Targeted Public Figures

Public figure	Ads detected	Detection window
Guillermo Tolosa — president, Banco Central del Uruguay	441	Jan 5 – Jul 21
Sergio Fogel — co-founder, dLocal	440	Mar 12 – Jun 25
Carlos Lecueder — businessman and real-estate developer	374	Jul 14 – Jul 22
Luis Lacalle Pou — former president of Uruguay	179	—
Juan Sartori — businessman and former senator	142	—
Álvaro Villar — physician, former director of Hospital Maciel	114	—
Roberto Canessa — cardiologist and Andes survivor	92	—
Luis Suárez — footballer	77	—
Enrique Baliño — businessman, founder of Xn Partners	73	—
Gabriel Oddone — minister of economy and finance	70	—

The top of the list is financial authority, not celebrity. The sitting central bank president, the co-founder of dLocal, and one of the country’s best-known developers take the first three places, with the economy minister also on the list. Luis Suárez, globally famous, ranks eighth with 77 ads against Tolosa’s 441. That ordering follows the product: 62% of these ads sell fake investments, and what makes an investment pitch land is financial credibility.

Lecueder’s 374 ads landed in eight days, July 14 to 22 — more than Lacalle Pou and Sartori combined. For scale, Tolosa’s 441 is roughly five times the 87-ad campaign impersonating Gustavo Petro that Teleamazonas aired in Ecuador, and nearly nine times the 51-ad David Vélez operation Forbes Colombia called a “machine.” Different windows and scan depths, so not a like-for-like ranking — but the order of magnitude is the point.

The ghost on the list: “Leandro Celeste”

Revelum also found 229 additional ads recycling the same video and image used in the Sartori ads, attributed to a different name — “Leandro Celeste,” who does not correspond to any identifiable real person. It was excluded from the ranking as an invented alias layered on existing material.

Per the report, that reuse of one deepfake under different identities is how the operators evade detection: change the name, keep the footage, and the ad reads as new.

---

What the Ads Sell: 62% Are Fake Investment Platforms

• 62% — fake financial investment schemes
• 19% — fake or “miracle” products
• 2% — fake online casinos and betting

The remaining share was not broken out. But the 62% explains the ranking above: you do not need a central bank president to sell a skincare product.

---

How the Scam Works: One Script, Ten Faces

Revelum analyzed the video transcripts and found a near-identical script repeated for every public figure, changing only the name and the peso figures:

• The format imitates journalism. Each video is staged as a news segment.
• The proof is local. Scripts cite supposed queues of sign-ups in Montevideo, Punta del Este, and Paysandú.
• The authority is forged. The pitch references documents presented as “certified” by the Banco Central, the Comisión de Valores Mobiliarios, or the Ministerio de Hacienda — none of which are actually involved.
• The entry price is calibrated. A minimum deposit of 7,000 to 11,000 Uruguayan pesos, or US$250, to “activate” the platform.
• The promised return is absurd. Between 12,500 and more than 100,000 pesos a day, “without risk” and “without experience.”
• The real harvest is the phone number. The final form asks only for a name and a number. A supposed “advisor” then calls to “confirm the registration” and request the deposit.

That last step matters most. The deepfake is a lead generator; the money leaves on a phone call, with a person applying pressure no ad could. It is the structure we lay out in the deepfake ad fraud playbook.

---

How Fast It Grew: From 33 Ads in January to 1,244 in June

Monthly detected volume went from 33 ads in January to 1,244 in June, with a brief dip in April. July had already reached 1,072 by the 22nd, in fewer days.

Our co-founder José Javier Cordero told El Observador why:

“Scammers are using images of public figures to promote various scams. Nobody is stopping them right now. The growth reflects the new low-cost technologies scammers are using and the success they are having with their scams — the more money they make, the more campaigns they launch.”

Generation cost has collapsed, and nothing is imposing a cost on the other side. Ads that are not detected are not reported; ads that are not reported are not removed; campaigns that are not removed fund the next batch.

---

How This Data Was Collected

The search ran on a lightweight detection algorithm designed for a first sweep, not exhaustive coverage. The figures are a representative sample, not a final count — the real total is higher, not lower.

Country attribution was inferred from mentions of “Uruguay,” “uruguayo,” or “uruguaya” in the ad text, the video transcript, or the ad card copy, because the underlying data has no explicit country field. Ads targeting Uruguay that never name the country would not have been captured.

---

What to Do If Your Face Is on That List

Assume it is already running. Tolosa’s ads span January 5 to July 21 — more than six months before this report. Lecueder’s produced 374 in eight days.

Coverage warns the public; it does not remove anything. Campaigns keep scaling for months after major press coverage, because media coverage barely dents a deepfake attack surface.

Only removal shrinks the numbers. Detection sizes the problem; sustained takedowns reduce it — see detection vs. removal and how the takedown process works.

And if this reaches you as a member of the public: no central bank, regulator, or ministry endorses private investment platforms through social media ads. A video of Guillermo Tolosa promising you 100,000 pesos a day without risk is a video of somebody else.

---

What Revelum Does

Revelum detects and removes deepfakes at scale, monitoring advertising and social platforms for synthetic media that impersonates executives, public figures, and financial institutions, then filing and tracking takedowns until the content, pages, and domains are gone. The data behind the El Observador report came from that same infrastructure.

Contact Revelum

We’ll assess your situation and tell you what we’re seeing, typically within 24 hours.

---

These findings were first published by El Observador on July 26, 2026, in a report by Juan Pablo De Marco, based on detection data from Revelum. See also our full index of media coverage of Revelum’s investigations.

Revelum is a deepfake detection and removal service operating globally, with a focus on the Americas and Europe. We protect executives, public figures, political leaders, and organizations from AI-generated disinformation and fraud.

Disclaimer: This article is intended for informational purposes only and does not constitute legal advice. Every situation is different, and we strongly recommend consulting a qualified legal professional for guidance specific to your circumstances. Revelum’s services are operational in nature and do not replace legal counsel.