New research raises concerns about the security of video conferencing platforms like Zoom. A study by researchers from UK universities reveals that hackers can potentially steal passwords by “listening” to the sounds of your typing with an accuracy rate of around 90%.
The study highlights the ability of artificial intelligence (AI) to identify specific keystrokes based on the sounds they produce. Microphones on laptops and smartphones can pick up these distinct typing patterns, making them vulnerable to exploitation.
Published in early August, the research points to the growing risk of sound-based cyberattacks as video conferencing use surges.
The researchers trained an AI model to recognize the unique sound associated with each keystroke. This involved recording the sounds of a MacBook Pro keyboard being pressed 25 times for each of the 36 keys. This data was then fed into the AI model, allowing it to learn the specific patterns.
The results were concerning. The AI model achieved a staggering 93% accuracy in deciphering passwords from keystrokes captured during Zoom calls on a MacBook Pro. Furthermore, the accuracy rate jumped to an even higher 95% when the keystrokes were recorded using an iPhone 13 mini.
These findings highlight the importance of cybersecurity vigilance, even during seemingly mundane activities like video calls. Users are advised to avoid typing sensitive information like passwords while using video conferencing platforms.
