Skip to content

Acoustic Typing Masker

The Acoustic Typing Masker is a physical defense against AI-driven acoustic keyloggers. Modern machine learning models can accurately reconstruct what you are typing simply by listening to the sound of your keystrokes through a compromised microphone (e.g., a phone on your desk or a smart speaker).

This tool floods the room with randomized, synthesized mechanical keystrokes, effectively destroying the signal-to-noise ratio required by these models.

Acoustic side-channel attacks rely on the minute differences in sound produced by different keys on a keyboard. By recording the audio of a typing session and analyzing the waveform, an AI model can map these sounds to specific keys with alarming accuracy.

The Masker operates on a simple principle: Signal Degradation via Chaos.

Instead of trying to silence your keyboard (which is difficult and often unergonomic), the Masker introduces a continuous stream of false keystrokes into the acoustic environment.

  • Adjustable WPM: Match the masking speed to your actual typing speed (30 - 150 WPM).
  • Rhythm Chaos: Humans don’t type at a perfectly consistent rate. The Masker introduces realistic hesitation, bursts, and pauses to prevent the AI from easily filtering out the masking noise as a predictable background pattern.
  • Multiple Switch Profiles: Choose from various mechanical switch profiles (e.g., Cherry MX Blue, Brown, Topre) to match your physical keyboard’s acoustic signature.

Deploy the Masker on a device near your physical keyboard. For maximum effectiveness, the audio should be played through high-quality speakers capable of reproducing the sharp transient sounds of mechanical switches.

Tip: If you use a mechanical keyboard, select the Masker profile that most closely matches your switches. If you use a membrane keyboard, use a loud mechanical profile to overpower the subtle sounds of your typing.