AME: A Multi-Type Contributor Attribution Framework in Generative AI Markets
2026-06-15 • Machine Learning
Machine LearningComputer Vision and Pattern Recognition
AI summaryⓘ
The authors studied how to fairly divide the value created by generative AI when many different contributors are involved, like data providers and model trainers. They identified key challenges including measuring each contributor's data value, connecting ownership rights, and making sure the process is trustworthy. To address these, they created the AME framework that combines these steps into one system. Their tests show AME aligns well with human judgments about data value while being efficient to run. This work lays groundwork for fair payment systems in generative AI data markets.
Generative AIData contribution valuationData rightsValue allocationMulti-stage collaborationFine-tuningAttributionTrustworthy executionRevenue allocationData markets
Authors
Yang Shi, Songwen Pei, Yang Gao, Bingxue Zhang
Abstract
Generative AI enables value creation through multi-stage collaboration among heterogeneous contributors, including training data, base models, fine-tuning behaviors, and prompts. However, how to fairly allocate the data value remains largely unexplored. This paper formulates multi-stage generative AI value allocation as a new research problem and identifies three core challenges: heterogeneous data contribution valuation, data rights mapping, and trustworthy execution. We propose AME (Attribution-Mapping-Execution) framework, a unified framework that integrates data contribution valuation, data rights mapping, and trustworthy execution into a single workflow. Experimental results demonstrate that AME framework achieves data value allocation outcomes more consistent with human reference judgments while maintaining low-cost trustworthy execution. Our work provides an initial foundation for value assessment and revenue allocation in generative AI data markets.