Attribute Inference from Interactive Targeted Ads

2026-06-13

AI Summary

This study addresses the privacy risks in interactive targeted advertising, where user engagement may inadvertently reveal sensitive attributes. The work presents the first formal model of attribute inference channels in this context and introduces a reproducible evaluation framework comprising synthetic populations, an activity semantic layer, and disclosure policies, accompanied by an open-source benchmark toolkit. Leveraging synthetic populations calibrated with real-world data, the authors conduct comprehensive simulations employing Bayesian inference, supervised learning, positive-unlabeled (PU) learning, and adaptive attack strategies. Experimental results across 160 activities demonstrate that Bayesian and supervised attacks achieve AUC scores of 0.64–0.65, while aggregated reporting effectively eliminates individual-level signals, and both type filtering and randomized disclosure significantly degrade inference performance.

Abstract

Targeted advertising systems can pair audiences selected by advertisers with ad units that expose visible user actions. When an interaction remains linked to the campaign that elicited it, the advertiser may receive an observation tied to a user rather than only an aggregate report. We model that channel as a noisy oracle for attribute inference. The model separates targeting predicates, exposure, interaction, and disclosure. These boundaries capture the gap between eligibility and delivery, and the gap between interaction and advertiser visibility. We build a reproducible benchmark using synthetic populations calibrated with public data, each with known sensitive labels. A generated campaign semantics layer provides topic variants and response priors. The simulator generates the ground truth, event traces, disclosed observations, and metrics. The evaluation compares Bayesian, supervised, positive and unlabeled, and adaptive attacks under common campaign and disclosure definitions. The final evaluation uses four topic variants, seven simulator seeds, and two interaction settings. Repeated campaigns with identity exposure produce measurable but bounded inference signal. At $160$ campaigns, Bayesian and supervised attacks reach about $0.64$ AUC in the main setting and about $0.65$ AUC in the higher interaction setting. Disclosure policy is the strongest control. Aggregate reporting removes the evaluated oracle input tied to users. Type filtering and randomized disclosure reduce the released signal. The result is a model, artifact, and defense evaluation method for privacy in interactive targeted advertising. The code is available at https://github.com/P-HOW/Interactive-Ad-Oracle.

Problem

• attribute inference
• targeted advertising
• privacy
• interactive ads
• user profiling

Innovation

• attribute inference
• targeted advertising
• privacy oracle
• synthetic benchmark
• disclosure policy