Abstract
In the modern world, surveillance is everywhere. Everyone, from kindergarten through adulthood, is monitored by algorithms on social media, security cameras in stores, video doorbell cameras on their daily walks, and license plate readers on their drive to work. Even systems that appear to benefit users, such as loyalty phone numbers for stores and healthcare apps that track personal metrics, are collecting data that users have no control over. Over time, each new layer of monitoring is accepted a little more easily than the last. The normalization of surveillance demands extreme caution: as our tolerance for surveillance expands, so does the risk that we stop considering the potential harms, and simply accept surveillance as an unavoidable part of our lives.
My capstone project addresses the problem of students having difficulty finding unoccupied study spaces at school. To solve this problem, our team developed an occupancy detection device using millimeter-wave radar to detect the presence of a human in a study room. We developed an algorithm to boost the accuracy of the radar sensor and connected multiple devices to a website to display the occupancy of multiple rooms, as well as display historical data of occupancy over the week. The system was designed as a student productivity tool, but any device that tracks where humans are in real time could easily be repurposed as a surveillance system or be abused by bad actors, as anyone with a school email can access the site.
My STS research uses Critical Discourse Analysis to examine how AI-powered weapon detection companies rhetorically justify the use of their technology, expanding surveillance in K-12 schools. These companies, as well as schools, justify the use of this potentially ineffective and harmful technology using rhetorical strategies that make their adoption appear both inevitable and morally non-negotiable. Critical discourse is applied because this analysis requires examining not just what AI vendors and schools say, but how the discourse fuels social action and shapes the conditions under which institutions perceive this technology as the only solution to school violence. The analysis reveals that two ideologies have been naturalized in the discourse. The first is that school shootings, a statistically rare event, are a perpetual, imminent threat that is only solvable with technology. The second is the moralization of skepticism: questioning this technology becomes the equivalent of not doing everything possible to protect children’s lives. Together, these ideologies are used to justify the use of AI weapon detection systems because schools and parents are now more concerned with the perception of safety over actually considering this technology’s effectiveness. Vendors use these rhetorical strategies to dominate the discourse: they promote the sale of their technology by eliminating the consideration of non-technological solutions, which actually address the root cause of school violence, from the discourse.
My STS research reveals how quickly the framing of a technology can override scrutiny of it, leading to an uncomfortable conclusion. My capstone project was built to solve a genuine problem, but if the project was sold to a school, what happens after the technology leaves our hands? It could be first introduced to students as a supportive tool, but then additionally repurposed as security surveillance with little modification. My STS research demonstrates how the normalization of a technology can make its adoption appear as just “common sense.” Students who have already accepted the occupancy detection system, as well as all the other surveillance in schools, are already primed to accept additional surveillance if it’s promised to protect them. Therefore, such a system is not ethical to deploy in schools without robust guardrails to protect it from being used as anything but an occupancy detection sensor.
