AI

AI Can Guess Your Password by Listening to You Type, Study Finds

16
Artificial Intelligence (AI)

A new study has found that artificial intelligence (AI) can guess passwords just by listening to the sound of typing on a keyboard. This raises serious concerns for streamers, remote workers, and anyone typing near a microphone.

How It Works

Researchers from Durham University, University of Surrey, and Royal Holloway trained an AI system to recognize what keys are being pressed by analyzing their sound. Every key makes a slightly different noise when pressed. The AI learned to match each sound with the correct key.

In tests, the AI achieved up to 95% accuracy using recordings from a smartphone microphone and 93% accuracy during Zoom calls. It could even guess typed passwords without seeing the screen.

Why This Is Dangerous

If you’re typing passwords or private messages during a live stream or video call, someone could record the sound and use AI to figure out what you typed. This type of attack is called an acoustic side-channel attack.

Built-in microphones in laptops, phones, or smart home devices can pick up these typing sounds. That means you could be at risk without even realizing it.

Who Is Most at Risk?

  • Live streamers
  • Remote workers
  • Anyone using video conferencing platforms like Zoom, Skype, or Teams

If these users type passwords or sensitive info while the mic is on, they could unknowingly expose their data.

How to Stay Safe

Experts recommend the following steps:

  • Mute your mic when typing passwords or sensitive data during calls
  • Use biometric logins like fingerprint or face recognition
  • Enable two-factor authentication (2FA)
  • Use password managers to avoid typing passwords
  • Create strong passwords using symbols, numbers, and mixed cases
  • Type using both hands (ten-finger typing) to make sound patterns harder to detect

A Bigger Trend in Cybersecurity

This is just one example of how AI is changing cybersecurity threats. Hackers are now using AI for deepfakes, smart phishing, and even malware that learns how to avoid detection.

As AI becomes more advanced, users need to be more cautious and update their security habits to stay protected.

Written by
Sazid Kabir

I've loved music and writing all my life. That's why I started this blog. In my spare time, I make music and run this blog for fellow music fans.

Related Articles

Meta AI Not Available
TechAI

Meta Will Let AI Handle Most Privacy Reviews for Instagram and WhatsApp

Meta is planning to use AI to handle most of its product...

Sundhar Pichai Google
AITech

Google CEO Sundar Pichai Challenges OpenAI and Perplexity Over AI Search Traffic

Google CEO Sundar Pichai has openly challenged competitors like OpenAI and Perplexity,...

Microsoft
TechAI

Microsoft’s Chief Product Officer Reassures Coders Amid Layoffs: AI Is Transforming, Not Replacing, Software Development

Microsoft’s Chief Product Officer, Aparna Chennapragada, recently addressed concerns surrounding the future...

DeepSeek Mobile
AITech

Powerful AI on a Budget: DeepSeek’s Distilled R1 Model Needs Just One GPU

DeepSeek, a Chinese AI lab, has released a new smaller version of...