A new cyber threat has come up with an artificial intelligence (AI) model that is reportedly capable of stealing one's system passwords by just "listening" to the keyboard sounds.
A new study by US-based Cornell University stated that the AI model, activated in a nearby smartphone, could reproduce a typed password in a laptop with 95 per cent accuracy.
According to the New York Post, the model was designed by a group of UK-based computer scientists. It was trained to recognise keystrokes which is now feared that could come handy for hackers.
Not just through a nearby smartphone, the AI tool proved to be extremely accurate in recognising the keystrokes through a laptop's microphone during a Zoom video conference.
"The ubiquity of keyboard acoustic emanations makes them not only a readily available attack vector but also prompts victims to underestimate (and therefore not try to hide) their output," the study said.
"For example, when typing a password, people will regularly hide their screen but will do little to obfuscate their keyboard's sound," it added.
In order to assess the AI model's accuracy, the researchers pressed on 36 keys on a laptop 25 times each, keeping the smartphone 17 cms away from the keyboard. They used varying pressure and finger during the process, and the model could recognise the distinguishing elements of each key press, including the sound wavelengths.
How to protect yourself from this risk
Researchers suggested some tips for users to protect themselves from this new cyber risk. One can have a mix of big and small letters and also characters for their passwords. It is likely to make it difficult for the AI tool to detect the password accurately.
When being on a video call, one can also think of having some noise near the laptop microphone.