Introduction
Data contributors are the cornerstone of the LazAI and Alith ecosystems. They can contribute privacy-sensitive data to earn rewards while retaining full control over how their data is used (e.g., for on-chain training, inference, or evaluation) though the Data Anchoring Token (DAT) and exercising governance rights with iDAO (Individual-centric DAO).
Data Anchoring Token (DAT) is a new semi-fungible token (SFT) standard to assetize your AI data specifically designed to tokenize AI datasets, models, and computation results with on-chain provenance, access control, and ownership rights.
What this hybrid-ability token standard DAT encodes:
Ownership Certificate: Proof of contribution or claim over datasets, models, or computation results.
Usage Right: Access quota to invoke AI services (e.g., model calls or agent execution).
Value Share: Economic entitlement to future revenue, proportional to the token’s value and share ratio.
Why AI Needs a New Token Standard
DAT vs NFTs and ERC-20s
Traditional NFTs can prove asset uniqueness, but they aren’t designed for modularity, usage tracking, or composable AI workflows. ERC-20s offer fluid transferability, but they’re blind to asset identity. Meanwhile, DAT is purpose-built for AI assets, a semi-fungible token that “transforms abstract data value into a tradable crypto asset while recording the full lifecycle of AI-related data.”
A quick comparison helps:
ERC-20 Tokens (Fungible): Interchangeable units like currency. Every token is identical and used for things like crypto balances or credit. These don’t point to any specific dataset or model – they just represent static value. Compared to an ERC-20 coin, a DAT is tethered to a specific asset class (e.g. a particular training dataset) and carries its identity on-chain.
ERC-721 NFTs (Non-Fungible): Unique tokens that tag a single asset (like a piece of art or a one-of-a-kind dataset). They prove ownership of that item but carry no flexible usage rules and can’t be split easily. Compared to an NFT, a DAT can incorporate usage quotas and revenue splits.
Semi-Fungible Tokens (Both): Combine aspects of both fungible and non-fungible tokens: they can start out identical and grouped (like fungible tokens), but also allow individual units to become unique later (like NFTs) and importantly, they can be split into smaller units or combined back together, depending on how they are used or needed.
In this way, DATs combine the best of both worlds for AI data: the indivisible tracking of NFTs, and the batch-handling/divisibility of fungible ERC tokens.