Designing Accountable and Private Systems

  • Date: Monday, April 20, 2026
  • Time: 2-3pm CT
  • Where: Morgridge Hall 6618 OR Zoom

Biography

Carolina Ortega Pérez is a computer science PhD student at Cornell Tech, where she is advised by Prof. Tom Ristenpart. Her research interests include applied cryptography, privacy, and tech abuse prevention.

Abstract

Historically, privacy-preserving technologies have prioritized security against malicious servers or external threats, as exemplified by the recent broad deployment of end-to-end encryption (E2EE) in messaging and cloud storage systems from companies such as Apple and Google. However, such threat models overlook interpersonal security adversaries, such as abusers in the setting of intimate partner violence. In turn, protocols neglect security goals such as accountability that provides mechanisms to cryptographically identify and verify misbehaviors, hindering progress towards a safer online ecosystem. 

In this talk, I will present my work on designing systems that provide both privacy and accountability for users. I will first talk about account security interfaces, which allow users to see their account activity. These interfaces enable users to detect whether a malicious user has logged into their account, but their entries are also unfortunately easy to spoof due to privacy concerns. To address this problem, we incorporate E2EE to include more trustworthy information while keeping it private from servers, and thus allow users to better detect whether account compromise has occurred. Then, I will talk about my work developing cryptographically secure message reporting mechanisms for interoperable E2EE messaging systems, where users from one platform can message users from other platforms. Finally I will discuss current research regarding offline finding systems (i.e., Bluetooth trackers).

Encrypted Access Logging for Online Accounts: Device Attributions without Device Tracking, USENIX Security 2025

Interoperable Symmetric Message Franking, ACM CCS 2025

SoK: Offline Finding Protocols for Lightweight Location Tracking, Preprint