Kalypta: AI Tool to Block Automatic Call Recorders
Deveillance has introduced Kalypta, a local AI tool that adds adversarial noise to audio streams to disrupt automatic speech recognition by services like Wispr and Granola.
Are you also annoyed by meeting recorders that are now being dragged into every call? Especially when they are connected without your knowledge via services like Wispr or Granola.
Developers from Deveillance have created a countermeasure — the product Kalypta.
How it works
A locally running LLM, adversarially trained against voice-to-text models, adds adversarial noise to your audio stream in real-time. This noise breaks the speech recognition on the other side, rendering automatic transcripts useless.
Downsides
Sometimes this noise makes your speech difficult to parse even for humans.
Outlook
Transcription model developers will have to regularly retrain their models to be robust against such attacks.
If the team manages to remove the obvious noise and polish the product, it will become a useful tool for maintaining call privacy.
Why it matters
AnalysisThis represents a significant development in the arms race between AI transcription services and privacy tools, offering users a way to protect confidentiality in corporate communications.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
Data Leakage Through Chatbots
The risk of compromising corporate secrets, access passwords, and confidential personal data through inadvertent transmission to public cloud chatbots (ChatGPT, Claude, Copilot). Learn how to protect accounts and disable model training on your queries.
Local LLM Inference
The practice of autonomously executing large language models directly on developer hardware (Apple Silicon, NVIDIA GPU) with guaranteed absolute privacy and zero dependency on the internet.
Part II. How to Work with Codex + Optimization/Security/Team
Codex is not just code autocomplete in a chat; it is a full-fledged engineering agent: it reads the repository, plans changes, edits files, runs tests, reviews its own diffs, and c...
How to Make AI Agents Safe: Stop Rules, Human Approval, and Sandboxes
An agent in production is a process that can write files, make API calls, change data, and consume tokens autonomously, without pausing and without your involvement at every step.