Vision Pro Typing Vulnerability: How GAZEploit Can Steal Your Passwords

Researchers discovered the GAZEploit attack, exploiting eye movements to predict keyboard inputs on Apple’s Vision Pro. By analyzing 3D avatars, they achieved up to 92.1% accuracy in lab tests, posing significant security risks. Apple patched this vulnerability in VisionOS 1.3, preventing Persona sharing during virtual…

Hot Take:

Who needs a crystal ball when you’ve got a Vision Pro? GAZEploit just made eye contact way more awkward.

Key Points:

  • GAZEploit attack tracks gaze patterns to predict typed keys on Vision Pro.
  • Uses recurrent neural networks trained on typing avatars for accuracy.
  • Research showed up to 92.1% accuracy in predicting typed letters.
  • Apple patched the vulnerability in VisionOS 1.3 after researchers reported it.
  • Potential for real-world exploitation during virtual meetings and typing tasks.

Membership Required

 You must be a member to access this content.

View Membership Levels
Already a member? Log in here