Researchers have created a system based on artificial intelligence (AI)which can determine which keys are pressed on a laptop keyboard with greater than 90% accuracy, based on sound recordings alone.

The research, published as part of the IEEE European Symposium on Security and Privacy Workshops, reveals how researchers used machine learning algorithms to create a system that could identify which keys were pressed on a laptop based on sound. an encryption device.

The study reports how researchers pressed each of the 36 keys on a MacBook Pro, including all letters and numbers, 25 times in a row with different fingers and varying pressure. The sounds were recorded during a Zoom call and on a smartphone placed a short distance from the keyboard.

The team then fed some of the data into a machine learning system that, over time, learned to recognize the characteristics of the acoustic signals associated with each key.

While it’s not clear which tracks the system used, Joshua Harrison, the study’s first author, of Durham University, said how close the keys were to the edge of the keyboard may have been a big influence.

“This positional information could be the main driver behind the different noises“, said.

The system was then tested with the rest of the data.

The results show that the system was able to accurately assign the correct key to a sound 95% of the time the recording was made via a phone call and 93% of the time the recording was made via a Zoom call.

The study isn’t the first to show that keystrokes can be identified by sound. However, the team says their study uses the most up-to-date methods and achieved the highest precision to date.

The researchers say the work is a proof-of-principle study and has not been used to crack passwords, which would involve correctly guessing sequences of keystrokes, or in real-world environments such as coffee shops.

However, they believe the work highlights the need for vigilance, pointing out that laptops – with their similar keyboards and widespread use in public places – are at high risk, and similar spying methods can be applied to any keyboard.