Buyer-Initiated Auction Mechanism for Data Redemption in Machine Unlearning
Bin Han, Di Feng, Jie Wang, Hans D. Schotten
TL;DR
This work tackles data redemption for machine unlearning under privacy concerns by introducing a buyer-initiated ascending auction where an AI service provider incrementally offers to buy data from users. The design jointly models the server's unlearning cost $C(y)$ and each user's privacy utility $U_i(y_i)$, and develops an incentive analysis showing how prices and data transfers depend on non-linear cost/utility curves. The proposed multi-round auction, along with four oversupply strategies and a post-auction discontinuity handling mechanism, achieves higher social welfare than static baselines such as GDPR, DNR, and single-price schemes, as demonstrated through extensive simulations. The approach enables flexible, prior-free data pricing that adapts to heterogeneous privacy preferences, reducing unlearning costs while preserving model quality in practical deployments.
Abstract
The rapid growth of artificial intelligence (AI) has raised privacy concerns over user data, leading to regulations like the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA). With the essential toolbox provided by machine unlearning, AI service providers are now able to remove user data from their trained models as well as the training datasets, so as to comply with such regulations. However, extensive data redemption can be costly and degrade model accuracy. To balance the cost of unlearning and the privacy protection, we propose a buyer-initiated auction mechanism for data redemption, enabling the service provider to purchase data from willing users with appropriate compensation. This approach does not require the server to have any a priori knowledge about the users' privacy preference, and provides an efficient solution for maximizing the social welfare in the investigated problem.