Norms and Attitudes about AI Surveillance: Evidence from a Survey Experiment

• In Event: Survey and Evaluation of Institutionalized Collaboration

Abstract

As advances in artificial intelligence technology accelerate, governments have struggled to keep pace with developing rules and procedures to protect rights and balance societal-level tradeoffs surrounding the new technology. This is especially true in the area of AI-enhanced surveillance. As surveillance technologies become more powerful, what determines how citizens and governments understand appropriate safeguards? This paper presents results of a pre-registered survey experiment that randomizes exposure to seven commonly cited issues surrounding AI surveillance technology: (1) its compatibility with international human rights obligations, (2) the possibility of limited use leading to more expansive and intrusive practices through “mission creep,” (3) the promise of using AI surveillance to combat crime and terrorism, (4) the concern that widespread adoption of Chinese produced AI surveillance technology may facilitate intelligence gathering on the part of the Chinese government, (5) the possibility that enhanced surveillance will be used to monitor civil society, dissidents, and journalists, (6) questionable reliability, including the potential for racial and other forms of bias, and (7) overregulation of this new technology may stifle innovation. Respondents are then asked about their overall concern about and trust of the technology, attitudes about regulation, and willingness to engage in political participation to advocate for government regulation of new surveillance technologies.

Authors

• Terrence Chapman, University of Texas at Austin

• Nivedita Jhunjhunwala, University of Texas Austin

• Huimin Li, University of Texas at Austin