A recent study from researchers at University of California San Diego is the first detailed assessment of companies offering school-based online surveillance services such as social media monitoring, student communications monitoring and online activity monitoring to middle and high schools. Schools pay for the services directly or may request federal grant funding to cover the costs.
Originally intended to support students’ mental health and prevent adverse school events, such as school shootings, the study found that school-based online surveillance companies are extensively monitoring students’ digital behavior, both during and outside of school hours, often using artificial intelligence (AI) with little human oversight or transparency. The study did not delve into success metrics on the original intention of these companies.
The research team identified 14 companies actively marketing online surveillance services frequently beyond school-issued devices and outside of school premises, raising concerns about privacy, equity and oversight.
Key findings include:
- 86% of companies monitor students 24 hours per day and 7 days per week, not just during school hours.
- 71% use AI for automated flagging of “concerning activity;” only 43% use human review teams. The definition of “concerning activity” was not generally clearly defined by the companies but typically appeared to include topics such as suicidal ideation and violence.
- 93% monitor school-issued devices; 36% also claim to monitor student-owned phones and computers but did not clarify if they monitor all activity on students’ personal devices.
- 29% generate student “risk scores” based on online behavior, which can be viewed at the student, classroom or school level.
Companies cited a variety of ways they gain access to student digital activity, including browser plug-ins, API integrations and device software. Further, the study found that companies rarely disclose pricing or performance data, and public-facing information is often vague or incomplete. Many companies collect and flag sensitive data, including students’ private messages and search histories. Some provide dashboards to administrators, after-hours alert systems and even direct crisis intervention, such as contacting law enforcement. However, most companies offer little to no public information about their algorithms, potential error rates or bias mitigation strategies.
The authors suggest that further research is needed on the prevalence of monitoring adoption across U.S. schools as well as how teachers and school administrators respond to alerts provided by the monitoring.
The study was published on July 8, 2025, in Journal of Medical Internet Research and was led by Cinnamon S. Bloss, Ph.D., professor at the UC San Diego Herbert Wertheim School of Public Health and Human Longevity Science and the director of the Center for Empathy and Technology, situated within the UC San Diego Sanford Institute for Empathy and Compassion. Alison O’Daffer, student in the San Diego State University - UC San Diego Joint Doctoral Program in clinical psychology, is the first author of the study.
# # #
Journal
Journal of Medical Internet Research