Skip to content
Submit incident
AII-1001-82787
AllegedThe record is a lawsuit, complaint or accusation that has not been decided.

Tbilisi call center allegedly uses AI-driven scripts to defraud over 6,000 victims of $35 million

A Tbilisi-based call center allegedly defrauded over 6,000 people of $35 million using AI-generated deepfake videos of celebrities and cloned investment platforms between May 2022 and early 2025. The scammers used ChatGPT to create fake credentials and replicate legitimate company websites, collecting victim funds through fraudulent bank accounts and cryptocurrency wallets.

May 2022 Β· Tbilisi-based call center (self-identified as skameri) Β· Fraud, scams and manipulation
May 2022
AII-1001-827872 sources ↓
Quick facts
Category
Fraud, scams and manipulation
Source
theguardian.com
Date
May 2022
Drafted by
Raidu Incident Detection

What happened

An industrial-scale boiler room fraud operating from three offices in Tbilisi, Georgia allegedly defrauded at least 6,179 people of approximately $35 million between May 2022 and the date of reporting. The group, self-identified as the 'skameri' (Georgian for scammers), employed 85 call handlers and support staff.

The scammers used AI-generated deepfake videos of public figures including broadcaster Ben Fogle to promote fraudulent cryptocurrency and investment platforms such as AdmiralsFX and Golden Currencies. Victims were directed to clone websites and contacted by call handlers posing as investment advisers.

According to Exante, a brokerage firm, scammers created a fake clone of the Exante platform using AI tools to rapidly replicate the company's trading interface and website. The fake platform targeted US clients despite Exante not offering services in the United States. Scammers opened a JPMorgan Chase bank account using a US address and crypto wallets to collect victim funds.

Reported impact

Mark, a franchise manager from East Anglia, initially invested Β£250 but within one month handed over Β£27,000 (approximately one year's salary) after being deceived by a deepfake video of Ben Fogle promoting AdmiralsFX. When he discovered his losses, he pleaded with his supposed investment adviser for answers about his money.

The scam reportedly collected $35 million from 6,179 victims across multiple fraudulent investment and cryptocurrency platforms routed through an apparent money-laundering network.

Index analysis

The use of AI tools including ChatGPT's image generation capabilities enabled rapid creation of fake passport images and cloned website interfaces, according to Exante's Head of Compliance. The scammers leveraged deepfake technology to impersonate credible public figures in promotional videos shared on social media platforms including Facebook and Instagram.

The Financial Conduct Authority had previously warned that AdmiralsFX and Golden Currencies were fraudulent clones created to scam people, but the warnings apparently did not reach all potential victims.

Governance controls

Exante did not report the incident to JPMorgan Chase as Exante has no presence in the United States and does not bank with JPMorgan. JPMorgan Chase stated it was investigating the fraudulent account.

The call centre allegedly continued operating despite being identified in media reporting as of March 2025, suggesting insufficient law enforcement intervention or regulatory oversight.

Classification

Category
Fraud, scams and manipulation
Industry
Financial and insurance services
Harm status
Actual harm
Intent
Intentional
Timing
Unknown
Harm types
Economic or property
Harmed
Consumers
Failure modes
Deepfake video
Consequences
None recorded
AI systems
ChatGPT, Deepfake generation tools

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

theguardian.com
Tbilisi call center allegedly uses AI-driven scripts to defraud over 6,000 victims of $35 million
financemagnates.com
Tbilisi call center allegedly uses AI-driven scripts to defraud over 6,000 victims of $35 million
Apr 2025AI-Generated Clone of Exante Brokerage Used to Defraud U.S. InvestorUnder investigationFraud, scams and manipulationfinancemagnates.comJul 2026JPMorgan CEO warns Anthropic's Mythos AI poses cybersecurity risks if widely availableWell documentedSecurity and cyberattacksQuartzCinco DΓ­asBusiness Insider+3May 2026US Regulators Pause Bank Cybersecurity Exams Over AI Model RisksUnder investigationSecurity and cyberattacksBloomberg BusinessInvesting.com IndiaGovernment TechnologyMay 2026Anthropic's Mythos model triggers concerns over AI-enabled cyberattacksReportedSecurity and cyberattacksBusiness24Stiri pe surseApr 2026Anthropic's Mythos AI overwhelms companies with vulnerability patchesWell documentedSecurity and cyberattacksRecorded Futurestoryboard18.comIT Security News - cybersecurity, infosecurity news+1