Hackers Can Use Artificial Intelligence to Steal User Passwords with Near-Perfect Accuracy

According to a recent study, hackers could potentially use artificial intelligence (AI) tools to steal user passwords “listening” to their keystrokes. UK-based computer scientists trained an AI model to identify the sounds produced keystrokes on a popular off-the-shelf laptop. When the AI program was enabled on a near smartphone, it was able to accurately reproduce the typed password with a remarkable 95% accuracy. Even during a Zoom video conference, where the AI tool “listened” to typing through the laptop’s microphone, it achieved an accuracy of 93%, setting a new record for this type of cyberattack termed as an “acoustic side channel attack”.

The researchers warned that users are often unaware of the risk posed bad actors monitoring their typing to breach accounts. They refer to this type of attack as an “acoustic side channel attack”. The researchers highlighted that people tend to underestimate this risk as they focus more on hiding their screens while typing passwords but do little to obfuscate their keyboard’s sound.

To determine accuracy, the researchers pressed 36 of the laptop’s keys multiple times with varying pressure and finger placement. The AI model was trained to “listen” to different elements of each key press, such as sound wavelengths. The smartphone used in the experiment was placed 17 centimeters away from the keyboard.

The study was conducted Joshua Harrison of Durham University, Ehsan Toreini of the University of Surrey, and Maryam Mehrnezhad at Royal Holloway University of London. It sheds light on yet another risk associated with AI technology. Prominent figures like Sam Altman, Elon Musk, and others have already voiced concerns about the potential dangers posed AI without proper safeguards in place.

In conclusion, this study serves as a reminder that advancements in technology, such as AI, can be exploited hackers. As users, it is crucial to stay vigilant and take necessary precautions to protect our sensitive information and accounts from unauthorized access.

Sources:
– Study published Cornell University
– Image source: Getty Images/iStockphoto (https://www.gettyimages.com)