# Facebook's Advertising Platform: New Attack Vectors and the Need for Interventions

arXiv: https://arxiv.org/abs/1803.10099 · HTML: https://mlsanity.com/paper/1803.10099/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Mar 2018.

## Built on by (most cited first, 3 of 3)

- [A Stochastic Optimization Framework for Fair Risk Minimization](https://mlsanity.com/paper/2102.12586/) (2021)
- [Private Stochastic Optimization With Large Worst-Case Lipschitz Parameter](https://mlsanity.com/paper/2209.07403/) (2022)
- [Why am I Still Seeing This: Measuring the Effectiveness Of Ad Controls and Explanations in AI-Mediated Ad Targeting Systems](https://mlsanity.com/paper/2408.11910/) (2024)

## Abstract (arXiv)

Ad targeting is getting more powerful with introduction of new tools, such as Custom Audiences, behavioral targeting, and Audience Insights. Although this is beneficial for businesses as it enables people to receive more relevant advertising, the power of the tools has downsides. In this paper, we focus on three downsides: privacy violations, microtargeting (i.e., the ability to reach a specific individual or individuals without their explicit knowledge that they are the only ones an ad reaches) and ease of reaching marginalized groups. Using Facebook's ad system as a case study, we demonstrate the feasibility of such downsides. We then discuss Facebook's response to our responsible disclosures of the findings and call for additional policy, science, and engineering work to protect consumers in the rapidly evolving ecosystem of ad targeting.