How to Trick AI: Users' Strategies for Protecting Themselves from Automatic Personality Assessment

Völkel, Sarah Theres; Haeuslschmid, Renate; Werner, Anna; Hußmann, Heinrich; Butz, Andreas

doi:10.1145/3313831.3376877

Cited by 25 publications

(19 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, recent studies show that other researchers have used Juji to build various interview chatbots, which matches our focus on aiding the design of effective interview chatbots [37,61,65,66]. Building and deploying an interview chatbot on Juji is very similar to creating a survey on a popular survey platform like SurveyMonkey or Qualtrics.…”

Section: Supporting Interview Chatbotsmentioning

confidence: 60%

Designing Effective Interview Chatbots: Automatic Chatbot Profiling and Design Suggestion Generation for Chatbot Debugging

Han

Zhou

Turner³

et al. 2021

Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems

View full text Add to dashboard Cite

show abstract

Section: Supporting Interview Chatbotsmentioning

confidence: 60%

Designing Effective Interview Chatbots: Automatic Chatbot Profiling and Design Suggestion Generation for Chatbot Debugging

Han

Zhou

Turner³

et al. 2021

Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems

View full text Add to dashboard Cite

show abstract

“…More broadly, our line of inquiry also fits in with studies on applying adversarial machine learning for beneficial goals, such as [3,14,28,62]. In most adversarial machine learning research, the party performing adversarial modifications is often referred to as the adversary.…”

Section: Adversarial Examples and Face Recognitionmentioning

confidence: 74%

FoggySight: A Scheme for Facial Lookup Privacy

Evtimov

Sturmfels

Kohno

2021

Proceedings on Privacy Enhancing Technologies

View full text Add to dashboard Cite

Advances in deep learning algorithms have enabled better-than-human performance on face recognition tasks. In parallel, private companies have been scraping social media and other public websites that tie photos to identities and have built up large databases of labeled face images. Searches in these databases are now being offered as a service to law enforcement and others and carry a multitude of privacy risks for social media users. In this work, we tackle the problem of providing privacy from such face recognition systems. We propose and evaluate FoggySight, a solution that applies lessons learned from the adversarial examples literature to modify facial photos in a privacy-preserving manner before they are uploaded to social media. FoggySight’s core feature is a community protection strategy where users acting as protectors of privacy for others upload decoy photos generated by adversarial machine learning algorithms. We explore different settings for this scheme and find that it does enable protection of facial privacy – including against a facial recognition service with unknown internals.

show abstract

“…More broadly, our line of inquiry also fits in with studies on applying adversarial machine learning for beneficial goals, such as [3,13,26,60]. In most adversarial machine learning research, the party perform-ing adversarial modifications is often referred to as the adversary.…”

Section: Adversarial Examples and Face Recognitionmentioning

confidence: 76%

FoggySight: A Scheme for Facial Lookup Privacy

Evtimov¹,

Sturmfels²,

Kohno³

2020

Preprint

View full text Add to dashboard Cite

Advances in deep learning algorithms have enabled better-than-human performance on face recognition tasks. In parallel, private companies have been scraping social media and other public websites that tie photos to identities and have built up large databases of labeled face images. Searches in these databases are now being offered as a service to law enforcement and others and carry a multitude of privacy risks for social media users. In this work, we tackle the problem of providing privacy from such face recognition systems. We propose and evaluate FoggySight, a solution that applies lessons learned from the adversarial examples literature to modify facial photos in a privacy-preserving manner before they are uploaded to social media. FoggySight's core feature is a community protection strategy where users acting as protectors of privacy for others upload decoy photos generated by adversarial machine learning algorithms. We explore different settings for this scheme and find that it does enable protection of facial privacy -including against a facial recognition service with unknown internals.

show abstract

How to Trick AI: Users' Strategies for Protecting Themselves from Automatic Personality Assessment

Cited by 25 publications

References 77 publications

Designing Effective Interview Chatbots: Automatic Chatbot Profiling and Design Suggestion Generation for Chatbot Debugging

Designing Effective Interview Chatbots: Automatic Chatbot Profiling and Design Suggestion Generation for Chatbot Debugging

FoggySight: A Scheme for Facial Lookup Privacy

FoggySight: A Scheme for Facial Lookup Privacy

Contact Info

Product

Resources

About