Web
Analytics
DL Seminar | Improving Privacy and Safety for Survivors of Intimate Partner Violence
  • Digital Life Initiative

DL Seminar | Improving Privacy and Safety for Survivors of Intimate Partner Violence



By Zicong Wei and Wenjia Zhang


On April 9th, 2020, the Digital Life Initiative welcomed Diana Freed, PhD Student in Computing and Information Science at Cornell Tech, presenting her work on “Improving the Privacy and Safety for Survivors of Intimate Partner Violence”. In her talk, Freed gave background on Intimate Partner Violence, explained the role of digital technologies in the IPV ecosystem, how intimate partners exploit technology, what types of spyware are used, how the IPV clinic helps end the technology abuse, and last, the transition to a remote clinic during COVID-19.


Digital Technologies & IPV


Intimate partner violence is a form of domestic violence with people involved in an intimate relationship. According to the CDC, the term intimate partner violence describes physical violence, sexual violence, stalking, or psychological harm by a current or former partner or spouse. Freed's research involves interviewing survivors of intimate partner violence and analyzing clinical computer security interventions from these clients.


Freed states that very little is known about technology abuse, and how it builds on other abuse-related issues such as trauma, safety, financial problems, violence addiction, and mental health issues. However, as technology has become ubiquitous in our daily lives, abuse via technology has become similarly common.


How Abusers Exploit Technology


As a result of her research on IPV, Freed has identified 4 common categories technology-related abuse. One is ownership-based attacks, in which an abuser has control over devices and accounts. A second is account compromise, where the attacker has access to the victim's account or device information without the victim's awareness. The third is harmful messages and posts. Finally, attacks can also expose of private information.


While these attacks do not seem technically sophisticated, they succeed because victims' threat models do not match the threats. For example, harmful posts can not be detected by standard approaches. Since the situations are complex and tech abuse is widespread, Freed thinks the solution to these problems should be taken into account in full social-technical context.

Spyware in IPV


Abuse via spyware is also a critical issue in IPV. Spyware apps can be downloaded to victims' devices by their abusers, and often go undetected. Despite its frequent use by abusers, the “state-of-art” for spyware detection is poor. Before Freed's research group looked into this field, there was no prior work investigating IPV spyware. The team built a machine learning technique to identify the potential of an app to be used for IPV, and found that Google and Apple’s app stores have abundant resources for spying on victims. The researchers created a taxonomy these apps, which includes personal tracking, mutual tracking, and subordinate tracking. As a result of this study, enough awareness of these apps was raised that companies such as Google took steps to restrict ad-serving on IPV-related search terms and take actions against violating apps and videos.


Ending Technology Abuse


As part of their efforts to end technology abuse, Freed and the IPV clinic team designed a clinical approach to help victims strengthen their computer security via having in-person consultations. While discussing with clients, tech consultants use IPV spyware scanning tools and examine clients’ iOS or Android devices. They try to understand clients’ situations, investigate possible tech vulnerabilities, and advise clients about potential next steps.


During the consultation study, the research team analyzed audio-recorded transcripts and notes from client sessions. They found that most clients first did not have a clear understanding of terms, and expressed their concerns using “client language”. They defined these terminologies and found a pattern of multiple concerns among clients - for example, clients report worrying about “calls from an unknown number”, identified by the researchers as call spoofing. In conversation with clients, consultants also use Technology Assessment Questionnaires to surface security vulnerabilities and a "technograph" to map clients' digital footprints. As for clients’ reactions, they find consultations valuable and some want to continue engagement.


This research has implications for technology design. Platforms can provide better control over and visibility into how accounts and devices are being accessed and used. Platforms can also improve notification about potential risks, including new or continued device logins. We need new tools to help users build better models of information flows and authorization dependencies.


IPV Concerns around COVID-19


Finally, Freed described how the IPV clinic is transitioning to a remote model as a result of COVID-19. She noted that reports of IPV have increased since the implementation of COVID-19 related restrictions, and that abusers may use COVID-19 to exert power over their victims. This means that even more people are seeking IPV related assistance - Freed's data shows an increasing number of people visiting domestic violence websites in March. Cornell Tech’s Clinic to End Tech Abuse started to do remote consults on March 27, but the remote model also poses challenges for both consultants and survivors.



Zicong Wei and Wenjia Zhang are MS students at Cornell Tech.


Contact

Address

Cornell Tech

2 W Loop Rd,

New York, NY 10044

Get Here >

DLI Queries

Jessie G. Taft

jgt43@cornell.edu