Generative AI Contribution Attribution Using Meta-Feature Correlation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing generative AI models lack a systematic method to attribute contributions of copyright-protected assets to their training, leading to unclear recognition and compensation for asset owners.

Innovation Solution

A method and system to determine contribution distribution of assets to a generative AI model by extracting meta features from assets, processing these features to calculate contribution scores, and outputting data for attribution and compensation based on these scores.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If generative AI models are trained using large datasets of copyrighted assets, then the model's generation capability and creativity are improved, but clear attribution and compensation mechanisms for asset owners are lacking

Engineering Contradiction:
Improvegeneration capabilityVSAvoidattribution information
Core Design Contradiction:
Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent applies preliminary action by embedding identification information (such as watermarks or metadata) into training assets before they are used to train the generative AI model. This allows the system to later trace and attribute generated content back to specific source assets and their owners, resolving the attribution loss problem while maintaining generation capability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback mechanisms where the system continuously tracks which training assets contribute to generated outputs and uses this information to provide attribution and compensation feedback to asset owners. This creates a closed-loop system that maintains both generation quality and proper credit distribution.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If contribution distribution of asset owners is determined through systematic analysis, then recognition and compensation for asset owners is improved, but the complexity of the system increases

Engineering Contradiction:
Improvecontribution measurementVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces intermediary components such as identification modules, feature extraction systems, and matching algorithms that serve as mediators between training assets and generated content. These intermediaries systematically track contributions without requiring complex direct analysis between all asset pairs, reducing overall system complexity while improving measurement precision.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the contribution determination process into distinct modular components: asset identification, feature extraction, similarity matching, contribution calculation, and compensation distribution. This segmentation allows each component to be optimized independently and simplifies the overall system architecture while achieving precise contribution measurement.

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If meta features are extracted and processed to calculate contribution scores, then the accuracy of attribution is improved, but the processing time and computational resources increase

Engineering Contradiction:
Improveattribution accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by extracting and processing only the most relevant meta features from training assets rather than analyzing all possible features. This selective approach maintains high attribution accuracy while significantly reducing processing time and computational resource requirements compared to exhaustive feature analysis.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent changes parameters such as feature extraction depth, similarity threshold values, and processing batch sizes to optimize the balance between attribution accuracy and processing efficiency. By dynamically adjusting these parameters based on resource availability and precision requirements, the system achieves accurate attribution without excessive processing time.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4654113A1Determination of contribution distribution of assets owners to the training and the products of a generative ai model
Publication Date: 2025.11.26 LAYER DATA LTD
  • EP4654113A1 patent drawingFigure 1A
  • EP4654113A1 patent drawingFigure 1B
  • EP4654113A1 patent drawingFigure 1C

AI summary

The present disclosure provides a solution for determining the extent of contribution of each copyright-protected assets owner to the training of a generative AI model. Furthermore, the present disclosure provides a solution for identifying the specific assets that contributed the most for the generation of a specific generated product being generated by the generative AI model. This is performed based on correlation of metrics related to meta features composing the assets and the generated product. By determining the contribution distribution of the owners to the generative AI model and a specific generated product, the owners can be attributed with recognition for their contribution. The recognition can be manifested in many ways, for example in attribution of copyright or an allocation of an income that is received for the use of the generative AI model according to a certain financial model. Therefore, by the solution of the present disclosure, the use of assets protected by copyrights in the training of generative AI models can be standardized.