No matter how many methods you use to protect yourself from hackers, hackers always find some new way. Would you believe it if someone hacked a password through the sound of typing? Maybe you shouldn’t believe it.
But in a new report, information has come out that a hacker can hack a password from the sound of typing. You can shield the screen and keyboard from prying eyes while typing passwords on the computer, but what about the sound? According to a report by JDNet, hackers can use a special technique to guess your password from the sound of the keyboard.
We often worry that someone is secretly watching our passwords, but we never pay attention to whether they listen to our passwords. This method of hacking is called acoustic side channel attack. In this type of password hacking, the focus is on the sound coming out of your keyboard.
Hackers crack your system password by analyzing sound. For this they have to use an advanced tool, which informs you about the exact letters and numbers you type. A team of experts conducted an investigation to understand how big this threat is. They used MacBook Pro 16 inches, which is a powerful laptop. The researchers placed the iPhone 13 mini on top of a soft cloth at some distance from the laptop, so that the sound could be captured. Besides, they also used the recording function of the laptop.
After this, all this data is used to train an AI-based smart computer program, whose job is to observe the typing sounds. After completing the training, this AI program is tested. In testing, this program can easily detect which button is which sound. It has been found to be 95 percent accurate. However, researchers have also discovered a way to avoid this. So don’t worry. Researchers have found that you can easily cheat an AI program.
For this you have to change your typing style. Additionally, you can type using Shift, which will provide an extra layer of security. If you’re on a video call, you can add background noise that prevents the AI tool from properly decoding your typing sounds.
Discussion about this post