Artificial intelligence (AI) is becoming so advanced it is able to listen to the sound of keyboard keys being typed and identify a password with almost 100 percent accuracy, according to new research published by Cornell University.
Researchers programmed an AI computer system to listen as a person typed a password on MacBook Pro keys over a phone and Zoom call.
The AI system began to recognize the individual sound of each key quickly.
Shortly after, the system was then able to identify the sound of each key with 95 percent accuracy over the phone and 93 percent over Zoom.
“When trained on keystrokes recorded by a nearby phone, the classifier achieved an accuracy of 95 percent, the highest accuracy seen without the use of a language model,” scientists write in the study.
Dr. Ehsan Toreini, from Surrey University’s Centre for Cyber Security, told The Times: “Each key has a unique audio, or voice, that can be fingerprinted to infer what is being pressed.”
The study has highlighted the potential risk of cyber attacks using the technology.
Co-author of the study, Dr. Ehsan Toreini, admitted: “I can only see the accuracy of such models, and such attacks, increasing.”
Professor Feng Hao of Warwick University echoed Toreini’s concerns, recommending that people no longer type “sensitive messages, including passwords, on a keyboard during a Zoom call.”
“Besides the sound, the visual images about the subtle movements of the shoulder and wrist can also reveal side-channel information about the keys being typed on the keyboard even though the keyboard is not visible from the camera.”
As The Daily Mail reported:
A similar study was done on an Enigma machine some years ago, which found that if it had been possible to implant a microphone, a similar outcome could have been achieved, but only with 70 percent accuracy.
“It gives you a hint of the tremendous improvement that has happened in the past five years in terms of the accuracy of the models, which somehow elevated the accuracy from 70-ish percent to around perfect results,” Dr. Toreini added.
He suggested that Apple consider adding random noises into keystrokes to ward off ‘side channel’ attacks and that technology such as Zoom should compress audio.
Side channel attacks attempt to get signals from a device and use acoustics, power consumption, and electromagnetic waves.