AI model can decipher keystrokes with 95% accuracy

• A new AI tool can identify passwords by listening to the sounds of keystrokes.
• The tool can be used to steal passwords during Zoom calls.
• Users should be aware of the risks and take steps to protect their passwords.

A recently developed AI tool can accurately identify passwords based on the sounds of keystrokes captured during various applications, including Zoom sessions, with an accuracy rate of around 90%.

Recent research highlights the potential risk of cyberattacks when typing passwords during Zoom calls, as hackers can exploit this to steal passwords using an AI-based tool.

The study, carried out by researchers from Durham, Surrey, and Royal Holloway universities, demonstrated that artificial intelligence can distinguish distinct typing patterns by analyzing keystroke sounds.

Published on August 3, the research emphasizes the growing threat of sound-based cyberattacks due to the rising use of video conferencing platforms like Zoom and the widespread presence of devices with microphones.

The researchers successfully trained an AI model by recording keystroke sounds while pressing each key on a MacBook Pro 25 times. This sound data was then used to teach the AI model to recognize the unique sound pattern of each key.

The AI model exhibited an impressive 93% accuracy in deciphering keystrokes from recorded typing sounds on a MacBook via Zoom. The accuracy rate improved further to 95% when capturing keystrokes using an iPhone 13 mini.

To stay informed about current events, please like our Facebook page https://www.facebook.com/BOLUrduNews/.
Follow us on Twitter https://twitter.com/bolnewsurdu01 and stay updated with the latest news.
Subscribe to our YouTube channel https://bit.ly/3Tv8a3P to watch news from Pakistan and around the world.

[embepdost slug = “/snapchat-ai-glitch-raises-concerns-among-users/”]