Hacking has become more sophisticated, and researchers at the University of Glasgow in Scotland have discovered ThermoSecure, a system that can guess computer, ATM and smartphone passwords in seconds using thermal imaging cameras and artificial intelligence, and they warn that criminals could be developing similar technology.
ThermoSecure is a thermal and artificial intelligence system that uses thermal imaging and artificial intelligence to detect heat from people’s fingertips on recently used keyboards.
When a password is typed on a keyboard or mobile device, it analyzes the visible signs of heat left by the fingertips. Since brighter areas in a thermal image indicate areas that have recently been touched, the order in which letters, numbers and symbols have been entered can be determined.
Using machine learning and 1,500 thermal images of recently used QWERTY keyboards, it was taught to recognize heat signatures and then draw conclusions about potential passwords.
ThermoSecure revealed 86 per cent of passwords when a thermal image was captured within 20 seconds of entering it. Within 30 seconds, the success rate had dropped to 76 per cent, and within 60 seconds, it had dropped to 62 per cent.
According to the team, longer passwords meant more security. ThermoSecure can crack only 67 per cent of the 16-character passwords in 20 seconds, increasing its success rate to 82 per cent for passwords with 12 symbols, 93 per cent for passwords with eight symbols and 100 per cent for passwords with six symbols. The writing style also has an impact; slow typers are considered more secure than fast typers.
For the researchers, PBT plastic keyboards are more secure than ABT plastic keyboards, so longer passwords or alternative authentication methods such as fingerprint or facial recognition should be used.
The sources for this piece include an article in CTVNews.