A new AI tool is capable of identifying passwords from the sounds of keystrokes captured during Zoom sessions or any other application that has access to your microphone with an accuracy rate of around 90%.
Recent research has revealed that typing your password during a Zoom call can put you at risk of a cyber attack as hackers can now ‘hear’ your keystrokes to use an artificial intelligence (AI) tool. can steal your password while doing so.
Research has shown that artificial intelligence devices have the ability to identify specific keystrokes by listening to typing sounds.
Research by researchers at the Universities of Durham, Surrey, and Royal Holloway has shown that microphones can identify different typing patterns.
The paper, published on August 3, revealed that with the adoption of video conferencing platforms such as Zoom and the widespread availability of microphone-equipped devices, the risk of voice-based cyber attacks has increased.
The AI tool studied by the researchers showed the ability to accurately decipher text, including passwords, from keystroke sounds, achieving an accuracy level of over 90 percent.
The team of researchers came to this conclusion after training an AI model.
They achieved this by pressing each of the MacBook Pro’s 36 keys 25 times, as well as recording the sounds they make.
Next, they fed this acoustic data into an AI model, allowing it to accurately recognize the unique patterns associated with each key.
The researchers’ AI model exhibited a remarkable accuracy rate of 93 percent in interpreting keystrokes from MacBook typing sounds recorded by Zoom video conferencing software.
Furthermore, when keystrokes were captured using the iPhone 13 mini, the accuracy rate increased to 95%.
