TRUEOPEN
Trueopen.AI

An Open GPU Network

for Verifiable and Private AI Inference

← Back to chatTRUEOPEN / DOCUMENTS
← Back to Home

Cryptographic Data Protection and Identity Privacy

Privacy protection covers both user identity and inference content. Users authorize tasks with a public-key address and a digital signature, without disclosing their names or home addresses.

They can optionally encrypt inputs and inference results using AES-256-GCM authenticated encryption. Data keys are wrapped using ECIES public-key encryption and delivered to authorized recipients according to their task roles.

The project team, storage nodes, and other organizations cannot decrypt user content solely by virtue of their identity or administrative privileges.

01 / Content Encryption

Input and Output Encryption

The user’s client encrypts the input. The Worker encrypts and returns the result in chunks, which the client decrypts for display.

02 / Task Authorization

Key Delivery by Task Role

The protocol selects Workers and Verifiers using future randomness. The required keys are then delivered to the selected nodes. Workers and Verifiers need access to the plaintext relevant to their assigned computations.

03 / Ciphertext Storage

Ciphertext Storage and Forwarding

Builders store ciphertext and wrapped keys without holding the unencrypted data keys. Materials are retained for the period required by the task and any review, then deleted according to the retention rules.

Data Availability

Storing, Verifying, and Encrypting Task Data

Verification must continue even after the Worker leaves. Multiple Builders store task materials in the data availability (DA) layer. On-chain commitments allow participants to check that retrieved materials have not been altered.

How Task Data Is Stored and Protected

Distributed Storage
Builders encode task data into fragments and distribute them across storage nodes. During the retention period, they check for and repair missing fragments so the materials remain available for verification and dispute resolution.
Recovery and Integrity Checks
Once enough fragments are retrieved, the complete materials can be reconstructed and checked against the recorded data commitments to confirm that nothing has been replaced or corrupted.
Encryption and Authorization
The user’s client encrypts the input, and the Worker encrypts the inference result before returning it. Builders store and forward ciphertext, which the user’s client decrypts for display.
AN OPEN NETWORK FOR VERIFIABLE & PRIVATE AI INFERENCEBuilt in the open ↗
TRUEOPEN / RESOURCES