.png){kind=logo}
A Hong Kong employee at Arup was tricked into transferring HK$200 million via a deepfake video call.
Over 50% of finance professionals in the US and UK have faced deepfake scam attempts.
The FBI warns AI voice cloning is now used in phishing scams, driving potential fraud losses to $40 billion by 2027.
Corporate deepfake fraud refers to scams where criminals use AI to mimic real executives. The aim is to trick employees into transferring money or sharing sensitive data. Deepfakes are AI-generated videos or audio clips that look and sound convincing. For example, scammers can clone a CEO’s voice or splice a video of a manager. This allows them to issue fake payment orders.
Cybersecurity experts say even ‘just a few seconds of audio’ is enough to replicate someone’s voice. These scams exploit our trust in familiar voices and faces. They turn AI into a powerful weapon for financial fraud.
The scam begins with attackers collecting public media like podcasts, videos, or voicemails of the target. These are used to train AI models to generate synthetic voices or videos. Then, a finance or operations employee is contacted by this ‘deepfake’ executive. The fraudster, often on a video or voice call, urgently asks for a financial transaction.
The FBI confirms that criminals now use AI-generated audio to impersonate public figures or relatives. The scam may include fake invoices or a fake lawyer to add credibility. One fraud setup included a Zoom meeting with a deepfaked CFO. It was followed by a cloned voice email confirming the transfer request. These attacks feel so real that the victim complies.
In 2024, British engineering firm Arup lost HK$200 million (about £20m). An employee in Hong Kong joined a video call with what seemed like real executives. The staff member followed fake instructions and wired funds to five accounts. Later, it was discovered that the executives were AI-generated voices and images.
In another 2024 case, scammers created a WhatsApp account using WPP CEO Mark Read’s photo. They even cloned his voice for a fake Microsoft Teams call. Luckily, WPP detected the fraud in time.
In Asia, deepfake scams are growing. A UK-based company’s finance worker in Hong Kong was tricked into transferring HK$4 million (US$512,000) via WhatsApp. The fraudster used real video footage of the CFO and added an AI-generated voice.
In Singapore (April 2025), scammers nearly stole US$499,000 using a fake Zoom call with a CEO. Authorities recovered most of the funds. Back in 2020, a UAE bank manager authorized a $35 million transaction. The call came from a fake director created using AI voice cloning. As early as 2019, a German CEO’s voice was deepfaked to con €220,000 from a UK firm.
Also Read: Deepfakes vs. Digital Trust: Is Online Security at Risk?
A 2024 survey of US and UK finance pros showed over 50% faced deepfake scams. About 43% of them suffered financial loss. A massive 85% called it an ‘existential’ risk to their organization. Deloitte forecasts that US fraud costs could hit $40 billion by 2027 due to AI.
Ahmed Fessi of Medius says deepfake scams are now seen as easy and effective. The FBI warns of ‘vishing’ campaigns using AI voices to gain trust. Bitdefender also reports that even consumers are now being pressured using AI voice clips. What once seemed like science fiction is now a real and fast-growing crime.
Here are some cybersecurity tips for business:
Verify Unusual Requests: Never act on a single call or message. If a manager urgently asks for funds, confirm through a trusted channel.
Try Official Sources: Call back using their official number or send a separate email. The FBI advises using a second verification method. Create a secret phrase executives can use to prove their identity.
Use Multi-person Approvals: No single person should be allowed to transfer large sums alone.
Audit: Set a rule requiring at least two managers for high-value transfers. Audit all approvals and record them to deter fraud.
Cybersecurity Training: Train staff and test responses. Teach teams, especially in finance, HR, and ops, about deepfakes.
Awareness: Warn them about audio glitches or robotic video behavior. Conduct drills using fake scenarios. Remind employees that real executives won’t rush payments or demand secrecy.
Implement Tech Policies: Use multi-factor authentication and encrypted channels. Monitor for unusual activity with AI detection tools.
Use Tools: Some tools can now flag deepfakes in real time. Update vendor protocols to verify new accounts. KPMG says proper controls are essential.
Limit Public Exposure: Executives should avoid posting clear videos or voice clips.
Avoid interviews or speeches that provide clean data for AI training. Limit personal content on social platforms. The less material available, the harder it is to clone.
Cybersecurity experts say that even as AI gets stronger, human awareness is the best defense. Deepfake scams combine social engineering with phishing and AI. Stay alert, verify requests, and don’t give in to urgency. One pause can save millions.
Also Read: AI-Generated Realities: Deepfake Trends to Watch in 2025
