AI hacking steals passwords from keyboards | Inquirer Technology

AI hacking method steals passwords by listening to keyboards

11:46 AM August 08, 2023

British university researchers trained an artificial intelligence model that can steal data from keyboard keystrokes with more than 90% accuracy. In other words, their AI program can listen to your keyboard to understand what you’re typing. As a result, this AI hacking method can take sensitive information like passwords and bank account information while you’re online.

Cyberattacks become more sophisticated as technology advances, so we must learn the latest hacking methods. This deep-learning model came from an academic study but eventually became a new tool for malicious individuals. Understanding that today could help you protect yourself from hackers.

This article will discuss the latest AI hacking technique involving keyboards. Then, I will explain how you could protect yourself from this method and conventional cyberattacks.


How does this AI hacking method work?


This study came from Durham University’s Joshua Harrison, the University of Surrey’s Ehsan Torcini, and Royal Holloway University’s Maryam Mehrnezhad. They published their findings on arXiv, and Bleeping Computer explained how it worked:

  1. The researchers started by recording keystrokes on a respondent’s keyboard.
  2. Then, they gathered training data by pressing and recording 36 keys on a MacBook Pro 25 times.
  3. Next, the experts produced spectrograms and waveforms from the recording to visualize differences for each key. In other words, they illustrated the keyboard sounds to distinguish between them clearly.
  4. Also, they used data processing to augment signals for identifying button presses.
  5. The researchers trained an image classifier from Google Brain called CoAtNet and experimented with learning, data splitting, and epoch parameters to boost prediction accuracy. These methods allowed the AI model to identify keystrokes.
  6. Then, they placed an iPhone 13 next to the computer to record keystrokes.
  7. Afterward, they listened to button presses via Zoom and Skype.

You may also like: OpenAI closes its AI detector due to low accuracy

The AI model achieved 95% accuracy from smartphone readings. Meanwhile, this artificial intelligence system got 93% accuracy from Zoom and 91.7% from Skype.

This Ai hacking method could become especially useful against people who use mechanical keyboards. These typing machines produce audible clicks when pressing buttons, making it easier for this deep-learning model to “hear” your keystrokes.

Silent keyboard users aren’t safe either because it also works on membrane-based ones. However, touch typing reduces its accuracy between 40% to 64%. Touch typing involves inputting words without looking at your keyboard.

How to defend against cyberattacks

Infographic highlighting 7 crucial steps for defending against cyberattacks, including network security, encryption, and more.

You can protect yourself from this new AI hacking type by learning touch typing. You can find numerous websites that teach that for free, but you need time to practice.

Also, you can use software that reproduces or filters keystroke sounds. Alternatively, playing white noise can muddle the AI program’s accuracy.

This technique is relatively new, so you’re unlikely to encounter hackers using it against you. However, you must defend yourself against conventional methods like phishing and scams.

You may also like: How to report a Facebook hack

Fortunately, existing techniques can protect you against the AI keyboard hacking technique and most cyberattacks. Stay safe online with these tips:

  1. Carefully handle your personal information by only sharing it with reputable platforms.
  2. Read a website’s terms and conditions before registering for it.
  3. Keep your mobile devices and computers updated with the latest firmware.
  4. Use strong passwords with capital letters, lowercase letters, numbers, and special characters.
  5. Enable two-factor authentication for your accounts. For example, you can set your Google account to require an SMS code before opening access.
  6. Avoid fake websites and emails. Don’t click anything if the platform or message looks strange or contains suspicious links.
  7. Secure your internet connection by placing a strong password.
  8. Avoid public Wi-Fi hotspots like airports and cafes. Hackers place fake hotspots to fool people into opening access to their devices. However, you can ask for an establishment’s official network if you need to use the Internet immediately.
  9. Use a password manager to store and input your security codes. As a result, you’re unlikely to share your passwords with strangers or keep them in unsafe places.
  10. Create backup files to protect your important files in case of a cyberattack. However, put your duplicates on an offline device not connected to your computer, such as an external hard drive.


British researchers developed an AI hacking method that takes data from keystrokes. It can listen while you type to get your passwords and sensitive information.

Your subscription could not be saved. Please try again.
Your subscription has been successful.

Subscribe to our daily newsletter

By providing an email address. I agree to the Terms of Use and acknowledge that I have read the Privacy Policy.

Protect yourself with the recommended methods above. Also, learn about the latest trending hacking methods to defend yourself immediately.

Start by reading more Inquirer Tech articles. The website reports the latest in artificial intelligence, cybersecurity, virtual reality, and more.

TOPICS: AI, Cybersecurity, interesting topics, Trending
TAGS: AI, Cybersecurity, interesting topics, Trending

© Copyright 1997-2024 | All Rights Reserved

We use cookies to ensure you get the best experience on our website. By continuing, you are agreeing to our use of cookies. To find out more, please click this link.