AI Scams Evolve With Digital Twins, Threatening Financial Ruin
Cyber expert warns criminals can use stolen data and voice clips to create convincing impersonations for fraud.
Sophisticated artificial intelligence is enabling a new wave of scams where criminals can create 'digital twins' of individuals, using harvested social media photos and voice snippets to impersonate victims. These advanced deepfake tools make it increasingly difficult for businesses to distinguish between a real person and a digital forgery, according to former Google executive Richard Mallam.
Mallam warned that this evolving threat could leave individuals unaware they are victims until they receive bills for debts incurred in their name. Criminals can compile leaked personal information with digital replicas of a person's face and voice to open financial products, lease vehicles, or take out loans and credit cards on their behalf.
Financial institutions are already experiencing a surge in fraudulent calls, with AI bots used to attempt account takeovers. Mallam advised against engaging with suspected scammers on the phone, as even brief voice recordings can be used to clone a person's voice. He noted that a small amount of personal data, combined with a voice clone, can be enough to bypass existing verification systems.
AI is not only increasing the sophistication of scams but also the scale at which they can be deployed. What was once a one-to-one attack can now be orchestrated by one person coordinating thousands of automated agents to target numerous individuals and businesses simultaneously. This trend is often fueled by organized crime networks, sometimes referred to as 'scam farms,' which recruit individuals to carry out mass fraud operations.
Mallam expressed concern that government services, such as the Australian Taxation Office (ATO) and Medicare, which hold vast amounts of sensitive data, will become increasingly attractive targets for cybercriminals. He suggested that without significant improvements in cybersecurity defenses, breaches of these critical services are almost inevitable.
To mitigate these risks, Mallam recommended several protective measures for individuals. These include using strong, unique passwords for every online account, enabling multi-factor authentication wherever possible, and adopting newer security methods like passkeys and biometric authentication. He also urged caution when sharing personal or financial information with unfamiliar online entities.
