LIVE
TRACECHAIN™FORENSICS
FRAUD INTELLIGENCE · CASE FILES · GLOBAL SCOREBOARD
REPORT FRAUD Find your country and who to contact ▸🔍 Search cases, companies, countries, agencies
INTEL
PIG BUTCHERING: now the #1 loss category in the US by estimated volume / HUSHPUPPI: the Instagram that became the evidence -- Case 001 open / BEC: one spoofed email, one wire, seven figures -- see the methodology / ROSS ULBRICHT: full pardon January 21 2025 after 11+ years -- Case 013 open / ALL FIGURES LABELED: estimate vs conviction vs market collapse -- see Methodology / UNODC 2023: $5.4B estimated annually from SE Asia pig-butchering compounds [UNVERIFIED] / PIG BUTCHERING: now the #1 loss category in the US by estimated volume / HUSHPUPPI: the Instagram that became the evidence -- Case 001 open / BEC: one spoofed email, one wire, seven figures -- see the methodology / ROSS ULBRICHT: full pardon January 21 2025 after 11+ years -- Case 013 open / ALL FIGURES LABELED: estimate vs conviction vs market collapse -- see Methodology / UNODC 2023: $5.4B estimated annually from SE Asia pig-butchering compounds [UNVERIFIED]
METHOD #50 OF 110

Deepfake Fraud

MODERATE-HIGH
SCALE (ENCYCLOPEDIA)
456% increase in AI scam reports
IN SHORT
AI-generated video/audio for impersonation. CEO deepfake: Hong Kong firm lost $25M to deepfake video call (2024). Voice cloning for phone scams. FaaS making accessible.
CONCENTRATED IN
Global (emerging)
WHERE IT STARTED
FIRST SEEN: 2019 (first reported voice-clone CEO fraud); mass adoption from 2023 · GOING FOR: About 7 years

The first reported financial deepfake was a 2019 voice-clone call that tricked a UK energy firm's CEO into a EUR 220,000 transfer (Wall Street Journal, 2019); the first large loss was Arup's $25.6 million in Hong Kong in January 2024.

The word 'deepfake' appeared on Reddit in 2017 for AI face-swapped video. The first reported financial use was in 2019, when criminals cloned a UK energy company CEO's voice to order a EUR 220,000 transfer. Cheap voice-cloning and real-time video tools after 2022 turned it into a mainstream fraud technique: in January 2024 a finance employee at Arup's Hong Kong office wired $25.6 million after a video call in which every other participant, including the 'CFO', was a deepfake. Politics followed (the January 2024 Biden robocall in New Hampshire), and deepfake romance and celebrity-investment scams now run at industrial scale. The FBI logged more than 22,000 AI-referencing complaints with $893 million in losses in 2025.

HOW IT WORKS

Attackers harvest a target's voice and face from earnings calls, conference videos and social media, then use cloning tools to impersonate them in a phone call, a video meeting or a pre-recorded clip. Corporate variants impersonate an executive to authorise urgent, confidential wire transfers (Arup, 15 transfers to five Hong Kong accounts). Consumer variants include 'family emergency' calls in a child's cloned voice, deepfake video calls in romance scams (a Hong Kong syndicate used face-swaps to pose as women), and celebrity endorsements of fake investment platforms. Political variants clone candidates to suppress or mislead voters.

Urgent, confidential money request from a senior figure by voice or videoVideo call where the other party's lip movements, lighting or blinking look slightly offA 'relative' in distress asking for money and refusing a call-back on a known numberCelebrity video promoting a crypto platform with a countdown or linkRefusal to verify identity by a second channel or a pre-agreed code word
HOW USEFUL IT IS TO CRIMINALS

Voice cloning now needs a few seconds of audio and a free or cheap tool, and real-time video swaps are within reach of small gangs; the Hong Kong syndicate was staffed by recent media graduates. Success depends on urgency rather than perfection, and the FBI counted $893 million in AI-related losses in 2025. Prosecutions are rare: the Arup attackers were never identified and the only US criminal deepfake trial of note (Kramer) ended in acquittal.

THE DAMAGE IT HAS CAUSED
$893 million
Losses in FBI complaints referencing AI, 2025 (deepfakes not isolated) — FBI IC3 Internet Crime Report 2025

The FBI attributed $893 million in 2025 losses to complaints mentioning AI (IC3, 2026), the best official figure; no body isolates deepfakes alone. The largest single case is Arup's $25.6 million (2024). Non-money harm: defamation (the Pikesville principal), election interference (the Biden robocall) and a collapse of trust in voice and video as proof of identity.

$893 million Losses in the 22,000+ complaints referencing AI received by the FBI in 2025 (FBI IC3 Internet Crime Report 2025)
$25.6 million Largest confirmed single deepfake corporate loss (Arup, Hong Kong) (CNN Business, May 2024)
$46 million Taken by the Hong Kong deepfake romance syndicate before 27 arrests (CNN, Oct 2024)
WHO USES IT MOST
MOST OF ALL: 🇭🇰 Hong Kong ▸
Site of the largest confirmed corporate deepfake loss (Arup, $25.6M) and the largest deepfake-romance bust (27 arrests, $46M), with Hong Kong police the main public source of case detail (CNN, 2024).
Hong Kong (Arup loss; romance syndicate)United States (FBI: $893M in AI-related complaint losses, 2025)United Kingdom (Arup headquarters; 2019 CEO voice clone)Spain (Brad Pitt impersonation arrests)
GOVERNMENTS LINKED TO IT
No government is documented as using this method.
NOTABLE PEOPLE WHO USED IT — RANKED BY PROFILE ORDER (NOT BY AMOUNT)
#1 Arup (victim) VICTIM; PERPETRATORS UNIDENTIFIEDREPORT PENDING
About $25.6 million (HK$200 million) · 2024
UK engineering group whose Hong Kong finance employee wired $25.6 million to fraudsters after a video call in which the 'CFO' and colleagues were all deepfakes; no arrests and no recovery reported.
SOURCE: CNN Business, 16 May 2024; CFO Dive
#2 Hong Kong deepfake romance syndicate (27 arrested) ARRESTEDREPORT PENDING
About $46 million (HK$360 million) · 2023-2024
Ring operating from a Hung Hom building that used face-swapping tech on video calls to pose as women and lure men across Asia into crypto investments; 27 arrested in October 2024.
SOURCE: CNN, 15 Oct 2024; The Record
#3 Steve Kramer CIVIL PENALTY; ACQUITTED OF CRIMINAL CHARGESREPORT PENDING
$6 million FCC fine · 2024
Political consultant behind the AI-cloned Biden robocall that told about 5,000 New Hampshire voters to skip the January 2024 primary; fined $6 million by the FCC, then acquitted by a state jury of voter-suppression charges in June 2025.
SOURCE: Bloomberg, 26 Sep 2024; NPR, May 2024; Wikipedia summary of June 2025 verdict
#4 Dazhon Darien CONVICTEDREPORT PENDING
No financial loss (reputational harm) · 2024
Former Pikesville High School athletic director who used AI to fabricate audio of the principal making racist remarks; pleaded guilty and was sentenced to four months in jail in April 2025.
SOURCE: The Daily Record; US News / AP, 29 April 2025
#5 Spanish 'Brad Pitt' gang (five arrested) ARRESTEDREPORT PENDING
EUR 325,000 (about $362,000) · 2024
Five people arrested in Spain for posing as Brad Pitt online, using AI-generated images, to persuade two women to 'invest' in his non-existent projects.
SOURCE: Time; Global News; RTE, Sep 2024
SOURCES
CNN Business, 'Arup revealed as victim of $25 million deepfake scam', 16 May 2024
CNN, 'Deepfake romance scam raked in $46 million from men across Asia', 15 Oct 2024
Bloomberg, 'FCC finalizes fine against Biden deepfake robocaller Steve Kramer', 26 Sep 2024
NPR, 'Criminal charges and FCC fines issued for deepfake Biden robocalls', 23 May 2024
US News / AP, 'Former school athletic director gets 4 months in jail in racist AI deepfake case', 29 Apr 2025
Time, 'Brad Pitt scam explained: police speak on arrests made', Sep 2024
FBI IC3 Internet Crime Report 2025
Wikipedia, '2024 New Hampshire Democratic presidential primary' (pointer to June 2025 acquittal)
🏠 HOME📁 CASES🚨 WANTED🎮 PLAY🌍 WORLD