Designing DSIC Mechanisms for Data Sharing in the Era of Large Language Models

2025-06-01

AI Summary

Large language model (LLM) training suffers from insufficient high-quality data due to legal constraints, privacy concerns, and heterogeneous costs faced by data providers; existing mechanisms either rely on unverifiable trust or ignore supply-side heterogeneity. Method: We propose the first data-sharing mechanism framework satisfying dominant-strategy incentive compatibility (DSIC), individual rationality, and weak budget balance. It integrates a quality-weighted marginal incentive auction (Q-MIA) with marginal utility tokens (MUT) into a hybrid mechanism, Mixed-MIA, augmented with differential privacy and verifiable execution. Contribution/Results: We prove Mixed-MIA’s DSIC property and budget feasibility. Experiments demonstrate substantial gains in high-quality data acquisition efficiency over volume-prioritizing and trust-based baselines, while ensuring robustness against misreporting and collusion, supporting privacy preservation and sustainable, trustworthy collaboration.

Technology Category

• Machine Learning: Privacy
• Multiagent Systems: Mechanism Design
• Game Theory and Economic Paradigms: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Data transparency and provenance
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Training large language models (LLMs) requires vast amounts of high-quality data from institutions that face legal, privacy, and strategic constraints. Existing data procurement methods often rely on unverifiable trust or ignore heterogeneous provider costs. We introduce a mechanism-design framework for truthful, trust-minimized data sharing that ensures dominant-strategy incentive compatibility (DSIC), individual rationality, and weak budget balance, while rewarding data based on both quality and learning utility. We formalize a model where providers privately know their data cost and quality, and value arises solely from the data's contribution to model performance. Based on this, we propose the Quality-Weighted Marginal-Incentive Auction (Q-MIA), which ranks providers using a virtual cost metric and uses Myerson-style payments to ensure DSIC and budget feasibility. To support settings with limited liquidity or long-term incentives, we introduce the Marginal Utility Token (MUT), which allocates future rights based on marginal contributions. We unify these in Mixed-MIA, a hybrid mechanism balancing upfront payments and deferred rewards. All mechanisms support verifiable, privacy-preserving implementation. Theoretically and empirically, they outperform volume-based and trust-based baselines, eliciting higher-quality data under budget constraints while remaining robust to misreporting and collusion. This establishes a principled foundation for sustainable and fair data markets for future LLMs.

Problem

• Ensuring truthful data sharing with legal and privacy constraints
• Addressing heterogeneous provider costs and data quality issues
• Designing incentive-compatible mechanisms for sustainable data markets

Innovation

• DSIC mechanism for truthful data sharing
• Quality-Weighted Marginal-Incentive Auction (Q-MIA)
• Marginal Utility Token for deferred rewards