{"id":3499,"date":"2026-03-10T09:11:56","date_gmt":"2026-03-10T09:11:56","guid":{"rendered":"https:\/\/alphasixtest2.com\/cs\/?p=3499"},"modified":"2026-03-10T09:51:15","modified_gmt":"2026-03-10T09:51:15","slug":"beyond-technology-mass-deepfake-guidelines-and-the-human-side-of-fraud-prevention","status":"publish","type":"post","link":"https:\/\/alphasixtest2.com\/cs\/beyond-technology-mass-deepfake-guidelines-and-the-human-side-of-fraud-prevention\/","title":{"rendered":"Beyond Technology: MAS\u2019s Deepfake Guidelines and the Human Side of Fraud Prevention"},"content":{"rendered":"

Introduction<\/h2>\n

When the Monetary Authority of Singapore (MAS) issued its September 2025 circular on deepfake risks, it marked a turning point for financial <\/span>institutions<\/b> and businesses. The guidance followed a series of documented incidents \u2014 from the <\/span>HK$200M Hong Kong CFO impersonation scam<\/b> (<\/span>The Guardian<\/span><\/a>) to a <\/span>\u20ac20M cryptocurrency fraud in Spain<\/b> (<\/span>PANews Lab<\/span><\/a>) and a <\/span>crypto scam in Hong Kong using AI-generated personas<\/b> (<\/span>Protos<\/span><\/a>). Even Singapore has seen close calls \u2014 including a <\/span>$499K near-miss prevented in March 2025<\/b> (<\/span>MAS<\/span><\/a>).<\/span><\/p>\n

The circular shows that MAS understands the scale of the problem. It does not just prescribe technology like liveness detection or endpoint detection tools. It emphasises the <\/span>human side of fraud prevention<\/b>: training <\/span>employees<\/b>, building stronger verification processes, and reinforcing separation of duties for high-risk <\/span>transactions<\/b>.<\/span><\/p>\n

For <\/span>companies<\/b>, the lesson is clear: fraud is not just a technical problem. It is a verification problem that spans people, process, and tools.<\/span><\/p>\n

The $67.4M Wake-Up Call<\/h2>\n

Hong Kong CFO Scam (HK$200M \/ $25M)<\/h3>\n

In February 2024, a Hong Kong employee was tricked into authorising HK$200M in <\/span>fund transfers<\/b> after attending a <\/span>video call where fraudsters used deepfakes of colleagues and the CFO<\/b> (<\/span>The Guardian<\/span><\/a>).<\/span><\/p>\n

The scam worked because the criminals controlled the conversation, eliminated doubt by having multiple \u201cparticipants,\u201d and applied pressure with fake urgency. The lack of a secondary <\/span>verification process<\/b> meant one employee\u2019s actions led directly to multimillion-dollar <\/span>losses<\/b>.<\/span><\/p>\n

Spanish Crypto Scam (\u20ac20M \/ $21M)<\/h3>\n

In April 2025, Spanish <\/span>authorities<\/b> investigated a deepfake-driven cryptocurrency fraud that cost investors \u20ac20M. Fraudsters used manipulated video <\/span>content<\/b> and fabricated endorsements to convince victims that the scheme was legitimate (<\/span>PANews Lab<\/span><\/a>).<\/span><\/p>\n

Unlike older scams, this case showed that <\/span>deepfakes<\/b> can be combined with polished marketing and credible <\/span>details<\/b> to bypass even cautious investors\u2019 judgement.<\/span><\/p>\n

Hong Kong \u201cSuperior Women\u201d Crypto Scam<\/h3>\n

In October 2024, Hong Kong police broke up a fraud ring that used <\/span>AI deepfakes of women<\/b> \u2014 marketed as \u201csuperior women\u201d \u2014 to lure victims into false relationships before convincing them to transfer money into crypto wallets. Losses reached tens of millions of dollars (<\/span>Protos<\/span><\/a>).<\/span><\/p>\n

This incident highlighted how <\/span>fraud tactics<\/b> prey on trust and personal connection, not just corporate processes. It demonstrated that both <\/span>executives<\/b> and ordinary <\/span>employees<\/b> can be victims of scams.<\/span><\/p>\n

Singapore Near-Miss ($499K)<\/h3>\n

In March 2025, MAS and police intervened to prevent a <\/span>$499,000 fraudulent transfer<\/b>, citing it as a near miss for Singaporean <\/span>businesses<\/b> (<\/span>MAS<\/span><\/a>). The incident showed that local verification processes can fail under pressure, but also that rapid coordination with <\/span>authorities<\/b> can stop the fraud in time.<\/span><\/p>\n

Together, these examples show why MAS has escalated its focus on corporate verification.<\/span><\/p>\n

What Is at Stake for Businesses?<\/h2>\n

For <\/span>companies<\/b> and financial institutions, the implications are wide-ranging:<\/span><\/p>\n

    \n
  1. Fraud is escalating<\/b> \u2014 attacks that once relied on crude phishing emails now use AI deepfakes.<\/span> <\/li>\n
  2. CEO fraud and business email compromise are evolving<\/b> \u2014 video impersonation is the new frontier.<\/span> <\/li>\n
  3. Verification failures undermine trust<\/b> \u2014 <\/span>employees<\/b> and <\/span>customers<\/b> alike struggle to distinguish authentic communications.<\/span> <\/li>\n<\/ol>\n

    The consequence is not only monetary <\/span>losses<\/b>. Regulatory fines, reputational harm, and loss of investor confidence can all follow when a <\/span>company<\/b> fails to address verification risks.<\/span><\/p>\n

    Why Corporate Verification Fails<\/h2>\n

    Biometric Authentication Defeat<\/h3>\n

    In August 2024, an Indonesian financial institution reported a <\/span>KYC bypass using deepfake photos<\/b> \u2014 proving that static <\/span>identity verification<\/b> is not enough (<\/span>MAS Circular, 2025<\/span><\/a>).<\/span><\/p>\n

    Social Engineering via Fake CEO Calls<\/h3>\n

    Fraudsters impersonate <\/span>executives<\/b> on video calls, pressuring <\/span>employees<\/b> to approve fraudulent <\/span>transactions<\/b>. The psychological manipulation is as powerful as the technical deception.<\/span><\/p>\n

    Trust Breakdown Inside Companies<\/h3>\n

    Employees<\/b> are conditioned to respect hierarchy. Without escalation <\/span>processes<\/b>, they may act on instructions from fraudulent <\/span>CEOs<\/b> without question.<\/span><\/p>\n

    Fraud Tactics MAS Has Highlighted<\/h2>\n

    MAS emphasised that <\/span>fraudsters<\/b> now combine <\/span>AI<\/b> with proven <\/span>fraud tactics<\/b>:<\/span><\/p>\n