PrivAI
  • About PrivAI
    • PrivAI’s Solution
  • Mission & Vision
  • Privacy & Automation Core
    • Dynamic Privacy Switching
    • Secure TEE Computation
    • Model Context Protocol (MCP) Bridge
      • Key Functions of the MCP Bridge
  • Ecosystem Features
    • AI Agent Marketplace
      • Create-to-Earn: Developer-Centric Model
      • Rent-to-Use: Permissionless Leasing for Users
      • Agent Discovery and Lifecycle
    • Cross-Chain Interoperability
      • Unified Execution Across Chains
      • Use Case Examples
    • Auditable Privacy Logs
  • Advantages
  • Technology
    • Trusted Execution Environments (TEE)
    • Model Context Protocol (MCP)
    • Agent Virtualization & Modular Deployment
  • Tokenomics
    • Token Allocation
    • Utility
  • Roadmap
  • FAQ
Powered by GitBook
On this page
  1. Privacy & Automation Core

Dynamic Privacy Switching

Dynamic Privacy Switching is the user-facing gateway to PrivAI’s core privacy architecture. It empowers users to manually or programmatically control the privacy level of each task they submit, offering flexibility based on the sensitivity, compliance, or performance needs of different operations.

When interacting with the PrivAI DApp, users can choose between two execution modes:

  • Classic Mode This is the standard, transparent execution path where tasks run publicly on-chain. It is ideal for non-sensitive operations such as open data queries, public smart contract interactions, or general-purpose AI tasks that do not require confidentiality.

  • Privacy Mode (TEE-Protected) In this mode, tasks are routed through Trusted Execution Environments (TEE) — secure hardware enclaves that isolate the computation from all external access, including from validators or node operators. These tasks are executed confidentially, and upon completion, they produce an on-chain cryptographic attestation, confirming the task was processed correctly and securely.

What makes this system “dynamic” is that users can switch modes on a per-task basis, allowing a single workflow to combine both public and private logic. For example, a DeFi agent can use Classic Mode to read token prices, then switch to Privacy Mode to calculate a large, undisclosed trade, ensuring execution logic and inputs are shielded.

This flexible switching capability is what gives PrivAI its granular privacy control, enabling personalized trust boundaries while preserving the efficiency and interoperability of public blockchains. Whether you’re a retail user, a DAO strategist, or an enterprise developer, Dynamic Privacy Switching lets you decide exactly what stays private — and what doesn’t.

PreviousPrivacy & Automation CoreNextSecure TEE Computation

Last updated 5 days ago