NFT-Based AI Model Output Verification System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack a mechanism to validate and trace the proof of ownership of decisions made by artificial intelligence (AI) and machine learning (ML) applications, which is essential for accountability and immutability.
Innovation Solution
An intelligent non-fungible token (NFT) based system that captures and stores AI-ML decision data and metadata in real-time or near-real-time, linking it to NFTs and storing this information in a distributed register for future validation, using a deep learning engine to extract and categorize data, and generating multiple NFTs for different categories of information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI-ML applications make decisions without ownership validation mechanism, then operational flexibility and speed are maintained, but accountability and traceability are lost
Solution Approach 1:
The patent introduces NFTs as intermediary objects that mediate between AI-ML decision-making processes and ownership validation. Each NFT encapsulates metadata about data usage, model version, and decision context, serving as a verifiable certificate that links decisions to their owners without requiring complex direct tracking mechanisms.
Solution Approach 2:
The patent creates digital copies of ownership information in the form of NFTs that can be independently verified and transferred. These NFT copies contain all necessary metadata to validate ownership and decision context, eliminating the need for complex centralized verification systems while maintaining accountability.
2Measurement precision
If real-time NFT generation and storage is implemented for every AI-ML decision, then ownership verification accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by generating NFTs at the moment decisions are made, capturing ownership and context metadata in real-time. This preliminary capture of verification data eliminates the need for subsequent retroactive verification processes, improving both accuracy and efficiency.
Solution Approach 2:
The patent extracts only the essential metadata needed for ownership verification into NFTs, separating critical verification data from the full decision-making process. This extraction focuses computational resources on capturing only the necessary ownership and context information, reducing processing overhead while maintaining verification accuracy.
3Loss of information
If comprehensive data and metadata are captured and stored for every AI-ML decision, then traceability and auditability are enhanced, but data storage requirements and system overhead increase
Solution Approach 1:
The patent extracts and captures only the essential metadata required for traceability and ownership verification, such as data usage rights, model version identifiers, and decision context. This selective extraction maintains comprehensive traceability information while minimizing storage requirements by excluding redundant data.
Solution Approach 2:
The patent transforms comprehensive decision data into condensed NFT parameters that preserve all necessary traceability information in a compact format. By changing the representation of ownership and context data into standardized NFT parameter structures, the system maintains full traceability capability while significantly reducing storage overhead.
Data Source
AI summary
The present invention is generally related to systems and methods for providing an improved authentication and verification system for artificial intelligence (AI) and machine learning (ML) model output. The invention immutably stores AI and ML decisioning data, resource data, and metadata in a non-fungible token format such that this data can be traced, relocated, and validated at a later time. Decisioning data of AI and ML model output is incorruptible and more reliable as a result.