The advancement and widespread availability of artificial intelligence (AI) technologies have given rise to new cybersecurity concerns, particularly the ease of executing deepfake attacks. The simplicity of these AI-driven tactics is now such that even young teenagers are capable of creating content that can be used maliciously.
Child’s Play Becomes a Cyber Threat
In the past, the widespread accessibility of malware like Trojan and Rat viruses led to a surge in digital crimes by minors. Today, the digital landscape has evolved to a point where even high school students can effortlessly fabricate content using the voices and images of friends or teachers, either as pranks or potential threats.
This development signifies a worrying trend, as the barrier to executing sophisticated attacks lowers. Consequently, there’s an increased risk of personalized attacks, including those targeting cryptocurrency investors, which are expected to become more common.
Deepfake Tech: A New Weapon for Fraudsters
IBM Security researchers have exposed a new attack method, termed “Audio-jacking,” which leverages generative AI technologies like OpenAI’s ChatGPT, Meta’s Llama-2, and voice deepfake tools. These systems can manipulate live conversations by altering spoken words to deceive listeners.
For instance, during a phone call, artificial intelligence can intercept and replace specific words or phrases, leading to potentially devastating consequences. An alarming case in Asia illustrated the danger when an employee was tricked into transferring $25 million following a fraudulent instruction that seemingly came from the company’s Finance Director.
IBM Security warns of the simplicity of creating these deepfakes, noting that with just a few seconds of voice recording, scammers can convincingly impersonate individuals and manipulate audio in real-time, posing serious threats to individuals and businesses alike.