mlsanityData to 7 Oct 2026

Data-Poisoning Audits for Causal Effect Estimation

arXiv 2607.19692 ↗First posted 22 Jul 2026stat.ML cs.LG stat.MEcited by 0

Not on the map. Why?

Abstract (arXiv)

Observational causal analyses increasingly pool records across sites, vendors, and collection systems, creating vulnerability to append-only attacks in which plausible records are strategically selected to alter a reported treatment effect. We develop a data-poisoning audit for augmented inverse-probability-weighted estimation. The analyst specifies a finite catalog of feasible records, an append budget, and nested source capacities, and the adversary selects a feasible subset to maximize movement in a prespecified direction. With preprocessing and nuisance fits held fixed, we propose a greedy scan that computes the exact finite-sample worst-case movement at every append budget. To account for nuisance refitting, we go on to derive a total-influence score combining each record's direct contribution with its effect through the propensity and outcome models. We further obtain a conservative finite-budget bound for the fully refitted estimate. Extensive simulations validate the exact result and show that total influence improves local refit prediction, while multisite and public-data analyses demonstrate material sensitivity at small append budgets. By translating adversarial data-composition risk into movement curves and critical budgets, the framework supports more reliable causal reporting and the design of source-level safeguards.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 39 references that are in the map. First posted on arXiv 22 Jul 2026.

Builds on

23 of 39 references.

References · 23

  1. 2017Understanding Black-box Predictions via Influence Functionscited by 1,714
  2. 2017Certified Defenses for Data Poisoning Attackscited by 288
  3. 2018Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learningcited by 219
  4. 2022Semiparametric doubly robust targeted double machine learning: a reviewcited by 129
  5. 2017Automated versus do-it-yourself methods for causal inference: Lessons learned from a data analysis competitioncited by 106
  6. 2016Balancing Covariates via Propensity Score Weightingcited by 65
  7. 2018Bounds on the conditional and average treatment effect with unobserved confounding factorscited by 39
  8. 2018Causal Inference with Noisy and Missing Covariates via Matrix Factorizationcited by 30
  9. 2019An Evaluation Toolkit to Guide Model Selection and Cohort Definition in Causal Inferencecited by 18
  10. 2018Robust Estimation of Causal Effects via High-Dimensional Covariate Balancing Propensity Scorecited by 12
  11. 2022Partial Identification with Noisy Covariates: A Robust Optimization Approachcited by 10
  12. 2024Stabilized Inverse Probability Weighting via Isotonic Calibrationcited by 8
11 more
  1. 2025Abstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacycited by 6
  2. 2024Data Poisoning Attacks on Off-Policy Policy Evaluation Methodscited by 4
  3. 2023Deception by Omission: Using Adversarial Missingness to Poison Causal Structure Learningcited by 3
  4. 2024Generalization Bounds for Causal Regression: Insights, Guarantees and Sensitivity Analysiscited by 3
  5. 2024Heavy Tail Robust Estimation and Inference for Average Treatment Effectscited by 3
  6. 2025Poisoning Bayesian Inference via Data Deletion and Replicationcited by 3
  7. 2021Outlier-Resistant Estimators for Average Treatment Effect in Causal Inferencecited by 2
  8. 2019Quantifying Error in the Presence of Confounders for Causal Inferencecited by 1
  9. 2021Covariate Balancing Methods for Randomized Controlled Trials Are Not Adversarially Robustcited by 1
  10. 2024Exploiting Missing Data Remediation Strategies using Adversarial Missingness Attackscited by 1
  11. 2025A Distributionally Robust Framework for Nuisance in Causal Effect Estimationcited by 1

Built on by

Not cited yet.