In Active Development

The Open Standard for
AI Permissions

A decentralized protocol for managing consent between humans and AI systems. No single entity controls it.

View on GitHub Read the Standards

Why LLMConsent?

Truly Decentralized

No admin keys, no central control. Built on blockchain with community governance.

User Sovereignty

Users own their consent tokens and control exactly how AI systems use their data.

Attribution Tracking

Cryptographic proofs showing how training data influenced model outputs.

Fair Compensation

Automatic micropayments when your data is used for training or inference.

Digital Twins

Persistent user models that evolve and can be shared across AI systems.

Agent Permissions

Granular control over what autonomous AI agents can do on your behalf.

The Four Core Standards

View the full technical specifications on GitHub

LCS-001

Core Consent

Basic consent tokens for training, inference, and synthetic data usage with attribution tracking.

LCS-002

Digital Twin

User-owned persistent models that evolve with interactions across AI systems.

LCS-003

Agent Permissions

How autonomous agents request and receive permissions with delegation chains.

LCS-004

Memory Sharing

Cross-agent memory pools for continuous experiences while maintaining control.

Simple Integration

# Grant consent for AI training
from llmconsent import ConsentClient

client = ConsentClient(network="arbitrum")

# User grants consent with automatic compensation
consent = await client.grant_consent(
    data="my_data.txt",
    purpose="training",
    models=["gpt-5", "claude-3"],
    compensation=0.001  # ETH per use
)