ERC-7007

ERC-7007: Inference assets

AI empowers anyone to become a creator, lowering barriers to entry and expanding the content market. However, AI-generated works lack copyright, raise ethical concerns over training data, and remain unverifiable, making it impossible to determine authorship or authenticity.

7007 Labs introduced inference assets to solve this. By leveraging onchain opML proofs, it establishes verifiable ownership, ensures transparency in model usage, and secures provenance for AI-generated content.

Built on the belief that AIGC is the future, 7007 Labs is creating a fully onchain AIGC ecosystem on inference assets, providing creators and developers with a protocol for asset issuance, verification, and fair monetization—ensuring AI-generated content is recognized, tradable, and equitably rewarded.

This innovative standard aims to enhance the efficiency, scalability, and security while maintaining compatibility with existing protocols like ERC-721 and ERC-1155.

Summary of Findings on IERC7007 and the 7007 Protocol

The interface defines essential functionalities for managing AI-generated content (AIGC) within the 7007 Protocol. Below are the key components and their implications:

2.1. Function Definitions

addAigcData Function: This function allows users to add AIGC data, ensuring that each token can carry specific information regarding its AI-generated content.

AigcData Event: This event is emitted whenever AIGC data is added, providing a transparent record of updates to the content.

verify Function: This function checks the validity of the combination of a prompt and associated AIGC data using opML (Optimistic Machine Learning) techniques. This verification process is crucial for ensuring that the generated content is authentic and aligns with the user's expectations.

2.2. Integration of opML

The 7007 Launch currently supports opML, which allows for efficient AI model inference on-chain. opML operates under a fraud-proof mechanism, enabling decentralized verification of AI-generated results while minimizing computational costs.

• The 7007 Launch currently supports opML, which allows for efficient AI model inference on-chain. opML operates under a fraud-proof mechanism, enabling decentralized verification of AI-generated results while minimizing computational costs.

In scenarios where an AI model behaves maliciously, the aigcData can be updated during a challenge period, allowing for corrections and ensuring that users receive accurate and trustworthy content.

• In scenarios where an AI model behaves maliciously, the aigcData can be updated during a challenge period, allowing for corrections and ensuring that users receive accurate and trustworthy content.

2.3. Update Extension

The integration of the optional Update extension allows for modifications to aigcData during the challenge period without compromising security. This feature ensures that any disputes or necessary updates can be handled effectively.

pragma solidity ^0.8.18;

/**
 * @dev Required interface of an ERC7007 compliant contract.
 * Note: the ERC-165 identifier for this interface is 0x702c55a6.
 */
interface IERC7007 is IERC165, IERC721 {
    /**
     * @dev Emitted when `tokenId` token's AIGC data is added.
     */
    event AigcData(
        uint256 indexed tokenId,
        bytes indexed prompt,
        bytes indexed aigcData,
        bytes proof
    );

    /**
     * @dev Add AIGC data to token at `tokenId` given `prompt`, `aigcData`, and `proof`.
     */
    function addAigcData(
        uint256 tokenId,
        bytes calldata prompt,
        bytes calldata aigcData,
        bytes calldata proof
    ) external;

    /**
     * @dev Verify the `prompt`, `aigcData`, and `proof`.
     */
    function verify(
        bytes calldata prompt,
        bytes calldata aigcData,
        bytes calldata proof
    ) external view returns (bool success);
}