SoundAuth — Secure, Seamless AI Agent Communication via Sound

AI authentication through advanced data-over-sound technology. Secure, frictionless, and ready for tomorrow's Agentic world

View Documentation
What is SoundAuth

SoundAuth is a protocol for secure authentication and encrypted communication for AI voice agents using advanced data-over-sound technology


As AI Voice agents become ubiquitous, they will interact with each other in the wild. However their current interaction is limited by natural language. Soundauth is a protocol based on MIT NANDA protocol which enables voice agents to authenticate each other and establish a secure and efficient communication link using data over sound.

In Action

Watch the demo below to see SoundAuth in action:

Key Features
Authentication

Agents exchange credentials verified via X.509 cryptographic certificates.

Validation can occurs via decentralized identity verification systems (NANDA’s distributed authentication layer).

Advanced Security

Robust protection against replay and impersonation attacks with proprietary sound-based protocols.


Upon identity confirmation, a secure session key is established using ephemeral keys for secure communication.

JSON based metadata

Agents can communicate their capabilities through JSON-based metadata, facilitating dynamic discovery and interaction among agents

Encrypted Communication

Supports secure, bidirectional communication between AI agents. Agents can continue to communicate using data over sound or switch to peer to peer web-socket or Https link.

How SoundAuth Works
Discovery

Voice agents discover each other.

Credential Exchange

Exchange X.509 certificates.

Verification

Verify identity via decentralized registry

Secure Access

Establish encryption key

Communication Communicate securely via the best-fit channel.


Use Cases and Applications
Phone-based Agent Authentication

Secure AI voice agent verification during calls using inaudible sound frequencies. Enables identity verification before exchanging sensitive information or processing transactions, all without human intervention.

Inter-Agent Negotiations

Secure communications for AI agents negotiating on behalf of users. Establishes trusted channels with cryptographic identity proof, protecting negotiations from impersonation attacks.

Secure Voice Transactions

Enhanced security for financial transactions between voice agents. Verifies identities before processing payments or sharing sensitive data, adding a layer of protection beyond traditional authentication.

Multi-Agent Collaboration

Security framework for AI voice agent teams. Enables trusted group formation and secure communication in real-time voice environments without disrupting conversation flow.

Developer Resources & Community
Explore Our Developer Tools

Jump to our Github Repo

Community Support

Join the Discord Channel

Join the SoundAuth Discord Server!

MIT NANDA Protocol


What is data over sound

Data over Sound transmits digital information via audio signals, either audible or inaudible (ultrasonic). This allows devices with basic speakers and microphones to exchange data without Wi-Fi, Bluetooth, or NFC.

Use Cases
  • Secure Authentication: Device verification through encoded sound signals.
  • Agent-to-Agent Communication: AI agents authenticating over voice calls.
  • IoT and Device Provisioning: Connects smart devices without manual pairing.
  • Payments and Access Control: Transmits encrypted credentials via sound.
Underlying Technology
  • Modulation & Demodulation: Data encoded into and decoded from sound waves
  • Encryption: 2-way encryption secures data.
  • Error Correction: Maintains accuracy in noisy environments.
Advantages
  • Efficient: Supports HTTPS, TCP/IP and UDP
  • Offline Friendly: Functions without internet connectivity.
  • Secure & Private: Difficult to intercept or manipulate.

Deploy SoundAuth for seamless, secure voice authentication in your environment.

Made with