Imagine typing your password on your computer and having AI listen to every keystroke and accurately guess what you are typing.
This is not a sci-fi scenario, but a real possibility, according to a new study by researchers at Cornell University.
They reported that AI can listen to the keystrokes you make on your computer and steal your passwords with up to 95% accuracy.
Let’s dive into this disturbing new information and see how you can protect yourself from this unfortunate situation.
How AI can eavesdrop on your keyboard
When the researchers tested an AI model listening to keystrokes, they trained the model on the intensity, waveform and time of each keystroke. The model listened for the keystrokes on a MacBook Pro and was able to repeat them with up to 95% accuracy.
However, having the model focus on the intensity, waveform and time of every keystroke means that AI will analyze the way you are typing and not how loud your keyboard sounds are while you’re typing.
There’s no need to rush out and get a brand-new keyboard that you think will sound quieter than the one you already have. The research team also tested this theory during a Zoom call and a Skype call, and each ended with the AI model reproducing keystrokes with over 90% accuracy.
This means that AI can potentially steal your passwords or other sensitive data by simply listening to your keystrokes, even if you are not showing your screen or keyboard during a video call. This is a serious security threat that could expose your online accounts to hackers.
How hackers can use your microphone to steal your passwords
A hacker can simply use this technique or method by installing malware on your devices with a microphone, such as your laptop or smartphone.
Once the malware is installed, it can gather data from your previous keystrokes and then feed them to an AI model, which will then use your microphone to listen in and repeat those same keystrokes to get your passwords.
Although this is a more complicated way for a hacker to get password information out of you, it’s not impossible, and you should still take precautionary measures to protect yourself from this potential threat.
How to keep yourself safe from AI replicating your keystrokes
Use strong and unique passwords
Use strong and unique passwords for your online accounts and change them periodically. A password manager can be a vital tool in preventing AI intrusions for two main reasons.
First, it consolidates all your passwords, automatically populating login fields, eliminating the need for you to recall or manually type them. This reduces the chance of AI detecting or predicting your keystrokes.
Second, password managers generate and safeguard intricate passwords for every account. This means that even if a hacker compromises one password, your other accounts remain secure and untouched.
Use 2-factor authentication
Another way to protect yourself from AI getting your passwords is by using 2-factor authentication. Having 2-factor authentication will add an extra shield that will prevent AI from getting into your account even if it guesses a keystroke correctly.
This usually comes in the form of having to provide extra information like confirming with a text message, email or even on a separate 2FA app such as Microsoft authenticator.
With 2FA in place, AI won’t be able to successfully steal your passwords with just a simple keystroke. Look into using 2-factor authentication on all your devices and accounts if it is available.
Have good antivirus software on all your devices
Keeping hackers and AI models out of your devices can be prevented if you have good antivirus software installed. Having antivirus software on your devices will make sure you are stopped from clicking on any potential malicious links that may install malware on your devices. That’s the first step a hacker might use to get that AI model to record your keystrokes.
See my expert review of the best antivirus protection for your Windows, Mac, Android and iOS devices by visiting Cyberguy.com/LockUpYourTech
Kurt’s key takeaways
Ultimately, getting a new keyboard is not the solution to avoiding AI models stealing your keystrokes.
The best you can do is follow all my suggestions above and make sure you’re regularly checking all your accounts to spot any suspicious activity.
This is a scary scenario; however, it can likely be avoided as long as you are vigilant and attentive to your information.
What more could AI companies like OpenAI be doing to make sure that hackers can’t use their models to assist them with malicious activity? Let us know by writing us at Cyberguy.com/Contact
For more of my security alerts, subscribe to my free CyberGuy Report Newsletter by heading to Cyberguy.com/Newsletter
Copyright 2023 CyberGuy.com. All rights reserved.