Future Finance Research Institute

What if AI Could Clone your Voice and Ask for Crypto?

Let's take a minute to reflect on the perfect cyber-crime. Where would you be? What would you be doing? How would the hacker reach you? Who is the person that you would help without hesitation? Would you be distracted? These are the types of questions that hackers are actively searching for answers. Now, put AI and Crypto transactions into the mix and you have the perfect storm for a tidal wave of cyber crimes.

This article will focus on AI, crypto and voice notes and how the combination of these three innovative technologies can be exploited with advanced personalized hacking that should be a cause for concern, a cause for alarm and a reason to increase cyber security measures across all devices.

How do we use Voice Notes today?

When we listen to voice notes on our various chat platforms they are mostly used to converse with friends, family and business partners. Voice notes aka audio messages have been steadily gathering momentum as a communication channel but they have been around awhile, with whatsapp first introducing voice messages in 2013. In 2022 they noted 7 billion daily voice messaging sent via whats app.

A study of voice notes in 2024 found some interesting personal reasons for choosing voice notes over text with an emotional connection being a key factor in how voice notes convey feelings in a better way. 61% of people agreed that tone is conveyed far better in voice notes than in texts, and 48% find them ideal for expressing complex ideas. They were also thought to strengthen personal bonds, with half of users reporting stronger relationships and 38% feeling less lonely thanks to this humanizing touch. In this sense voice notes are a trustworthy source of information, that is until AI can intercept this medium.

As voice-based technologies like voice notes and AI-powered voice assistants become more integral to our daily lives, they open the door to both convenience and heightened security risks. Security experts are urging the public to look beyond the vulnerabilities of voice notes alone and consider the broader threat posed by AI voice cloning.

Attackers today can use cloned voices to impersonate trusted figures, tricking victims into revealing private keys or recovery phrases. “The addition of realistic voice cloning to phishing scams drastically increases their success rate and makes them harder to detect,” warns Grace Dees, Cybersecurity Analyst of Resonance Security.

The Risks of AI-Generated Audio in Cryptocurrency

The increasing sophistication of AI-driven voice cloning tools, such as Descript, Respeecher, and iSpeech, poses a growing threat to digital security, particularly in the cryptocurrency world. Hackers can collect voice samples from publicly available sources, including voice notes, social media, or videos, and use them to bypass voice-based authentication systems.

As some wallets experiment with integrating voice authentication into multi-factor authentication (MFA) systems, hackers may use cloned voices to impersonate individuals and gain unauthorized access to sensitive accounts.

Speaking to Dees was an eye-opening discussion. She believes that organizations are not keeping up with the pace of change when it comes to adopting advanced fraud detection systems capable of identifying AI-generated audio. “Investing in these technologies is essential to countering synthetic voice attacks as they enhance organization’s abilities to distinguish between genuine and synthetic voices in real time,” Dees notes.

From her engagement with firms she also sees a pressing need for more awareness and education when it comes to the risks of AI generated audio. “Users must understand the risks of vishing and AI-generated audio, along with strategies to recognize and respond to these threats. Also, individuals can limit the amount of data available for malicious use by reducing their voice recordings on public platforms and practicing heightened skepticism toward voice-based interactions.”

Erosion of Trust leading to Delays in Adoption

Beyond cryptocurrency, AI voice cloning creates significant vulnerabilities in various systems, from smart home assistants to healthcare applications. Devices reliant on voice commands are especially susceptible to unauthorized access. A cloned voice could manipulate voice-controlled IoT devices to disable security systems, change settings, or cause disruptions in sensitive environments like hospitals.

The broader implications extend to trust. "The erosion of trust in voice technology could stall innovation," Dees explains, warning that users may hesitate to adopt these systems due to perceived insecurity. The difficulty of investigating and prosecuting audio-based attacks adds another layer of complexity, as attackers exploit legal and regulatory gaps to operate with impunity.

Voice is personal, voice is trustworthy, voice gives us a sense of connection but what if those voices could be cloned and used against us? How many of you wake up in the morning to voice notes from friends and family? How many of you send voice notes on a daily basis? Voice messages have become a key part of our everyday lives. Let’s consider how to use Voice, AI and Cryptocurrencies for good.