How student surveillance service Gaggle monitors online activities of about 5M US students for “suspicious or harmful content” using AI and human moderators
Caroline Haskins / BuzzFeed News : Tweets: @carolineha_ , @trekonomics , @mediasee , and @loisbeckett Tweets: Caroline Haskins / @carolineha_ : SCOOP: I got hundred of pages of documents + content moderation info for Gaggle, a student surveillance giant. It scans emails, calendar events, chats, documents, & everything students do. And it pays contract content moderators $10/hour to review it all https://www.buzzfeednews.com/ ... Manu Saadia / @trekonomics : They told us the internet was a force for good and instead we got mass surveillance of children's lives. https://twitter.com/... Neil Andersen / @mediasee : So many rich #medialiteracy discussion ops: Is this teaching Ss to accept surveillance as a fact of life? How is “Gaggle” ironic? What quality does $10/hour suggest? Tasks for the VP? How is the data secured/sold/shared with parents? https://www.buzzfeednews.com/ ... @A_M_L_ @zephoria Lois Beckett / @loisbeckett : Public schools are paying companies to scan everything their students write in emails and shared documents for signs of violence/self-harm. The humans who review those “red flags”? Some of them make $10/hr, no benefits. No background with kids required: https://www.buzzfeednews.com/ ...
Context & Ripple Effects
Caroline Haskins' document haul lands in the middle of a rush adoption cycle: as Education Week reported months earlier, US schools had been buying student-monitoring programs with little regard for effectiveness or privacy, and Gaggle is the first look inside what one of those contracts actually buys — AI plus a human review layer scanning emails, chats, documents, and calendar entries for roughly 5M students.
The reporting exposes two details that shape everything after it: contract moderators are paid around $10/hour with no child-specific background requirements, and the flagged categories extend well past violence into routine student writing. Both become recurring fault lines — Gaggle later drops LGBTQ keyword flagging after discrimination concerns, and by 2025 an AP investigation finds no research showing the tool increases safety at all.
First-order effects
- About 5M students' school-issued communications are being read by low-paid contract moderators earning ~$10/hour, with no benefits or child-specific screening requirements disclosed in the reporting.
Second-order effects
- Rival GoGuardian scales the same model — its usage explodes during pandemic lockdowns, adding to what critics describe as kids' psychological strain — turning student surveillance into a competitive category rather than one vendor's practice.
Third-order effects
- The pattern hardens into infrastructure: by 2025, districts run these tools over students' AI chatbot conversations for self-harm detection while Gaggle itself has grown to ~6M tracked students despite no published evidence the monitoring improves safety — leaving schools paying for surveillance whose efficacy was never established when they first adopted it.
The trend: Student-monitoring software is evolving from post-Columbine-style social media scanning into always-on AI surveillance of every school communication, expanding scope faster than any evidence base or privacy framework can constrain it.