Scam calls are not new — but the technology behind them is changing fast.

According to warnings from Swinburne University of Technology, the rise of AI-driven voice cloning is significantly increasing the risk and impact of phone-based scams. In Australia alone, reported losses from scam calls reached AUD $25.8 million in the first half of 2025, highlighting how quickly these techniques are spreading.

What makes this wave of scams different is not just the scale, but how convincing and personal they have become.

Why Voice Cloning Is So Effective

Dominique Carlon, an AI expert at Swinburne University of Technology, explains that scammers are now using generative AI to impersonate people across text, images, audio, and video. Among these, voice cloning stands out as particularly confronting.

A familiar voice can override instinctive caution. When a call sounds exactly like a child, partner, manager, or colleague — especially when paired with urgency — people are far more likely to act before verifying what is happening.

Unlike older synthetic voices that often sounded robotic or unnatural, modern voice-cloning systems can sound realistic with only a short sample. These samples are easy for scammers to obtain from social media videos, voicemail greetings, or publicly shared audio clips.

The result is a live interaction where emotional pressure builds quickly, leaving little time to pause, think, or verify identity.

How Voice Cloning Is Being Used in Scams

Carlon links AI-generated voice impersonation to a growing range of scam types. These include urgent requests for money or personal information, romance scams, job-related scams, and ransom-style calls designed to pressure victims into immediate action.

Because these calls unfold in real time, they compress manipulation and decision-making into a single moment. The combination of a trusted voice and a time-sensitive request makes it harder for people to apply the usual warning signs they may recognise in emails or messages.

The Psychology Behind the Technology

While AI provides the tools, Carlon stresses that the success of these scams still relies heavily on psychology.

Scam calls are designed to trigger instinctive human responses such as fear, urgency, hope, or trust. By creating emotionally charged scenarios, scammers push people to act before their rational thinking has a chance to engage.

AI amplifies this effect by allowing scams to be more personalised and more scalable than ever before. As these techniques evolve faster than public awareness, even people who consider themselves cautious can be caught off guard.

Who Is Most at Risk — and Why It Matters

Carlon notes that scam harm does not affect everyone equally. People under financial stress, those facing visa uncertainty, individuals with language barriers, or people who already mistrust institutions can face greater exposure and greater consequences.

Scams don’t just take money — they deepen inequality. Shame and embarrassment often prevent people from speaking up or seeking help, which allows scam operations to continue unchecked.

Slowing Down Is the Most Powerful Defence

Swinburne’s guidance focuses on a simple but effective framework: Stop. Check. Protect.

• Stop if you feel pressured to act quickly — urgency is a major warning sign
• Check by independently verifying the request, such as calling back using an official number
• Protect yourself immediately if information or money has been shared, by contacting your bank and reporting the incident

Carlon emphasises that focusing on what is being asked — rather than how convincing the voice sounds — is critical. Slowing down, even briefly, can break the scammer’s momentum.

Just as importantly, she encourages people to talk about scam experiences. Scams thrive in silence, and intelligence alone is not a protective factor when deception is this sophisticated.

A CSB Perspective

At CSB, we’re seeing a clear shift: scams are no longer just technical tricks — they are designed experiences that combine AI, psychology, and trust.

As AI-generated impersonation continues to evolve across voice, video, and text, awareness and verification habits become just as important as technology. Education, open conversation, and removing stigma are essential parts of reducing harm.

Anyone can be targeted by a sophisticated scam. Staying safe isn’t about being smarter — it’s about slowing down, verifying independently, and knowing what to do if something goes wrong.

Source & Expert Commentary
Insights and figures attributed to Swinburne University of Technology and Dominique Carlon.