Warn of thermal attacks, which can guess a password in seconds

Oct. 10 (Portaltic/EP) –

A group of scientists from the University of Glasgow has developed a system based on artificial intelligence that guess the password of a user in seconds by detecting the heat signature of fingers on the keyboard and screens.

The system, called ThermoSecure, reveals the risks of thermal attacks in cybersecurity, especially with the increasing use of artificial intelligence, such as the University in a Press release.

Password theft by a heat attack it requires no expert knowledge, just looking at thermal images of a computer strategically placed next to a screen or keyboard where the key characters have been entered.

The image will show in colors the keys that the user has pressed, being brighter those recently pressed. With a careful look you can guess a person’s password about 30 to 60 seconds after it has been entered.

The developed system is capable of identifying long passwords of 16 characters in 20 seconds with a rate of 67 percent of correct attempts. In shorter passwords, the success rate increases; it rises to 82 percent with 12-character keys and up to 93 percent with eight characters.

If passwords are reduced to six characters, the chance of success rises to 100 percentas stated in the press release.

The user’s typing style it also contributes, so that a person who types slowly and tends to rest their fingers longer on the surface of the keys will create a more durable heat signature.

It also influences the material the keyboard is made of, as it affects the ability to retain heat. Thus, the study indicates that keyboards made of ABS plastic reduce guessing success by half as many times, while in the case of PBT plastic, success is around 14 percent of attempts.