GAT Labs

GAT Labs — Browser-level student activity with an unusually strong data-egress dimension. GAT Shield operates inside Chrome and Chromebook management for districts, and captures browsing, downloads, chats and device location, plus category, URL and keyword policy verdicts. The genuinely scarce element is egress visibility: it monitors file uploads leaving the school domain to external sites, including AI tools, and runs real-time data-loss-prevention rules that detect or block sensitive-data transfers across all browser activity. Combined with organisational-unit level allow and deny lists, d…

GAT Labs offers (Alternative, Sentiment, Reference) — Browsing and audit volume is modest at 18 staff and a niche base — est. low millions of policy events per year. The valuable subset is much smaller and that is fine: file-upload egress events to external and AI destinations are sparse by nature, plausibly in the tens of thousands to low hundreds of thousands per year, and their value is per-event rather than statistical. Do not expect a large corpus here; expect a dense, precisely-labelled narrow one..

The egress view is the differentiated use: a ground-truth map of which consumer AI tools schools and students actually push files to, which is a data-protection and vendor-management question that AI labs, edtech compliance teams and district CISOs cannot currently answer and that no screen-time product captures. Also: training and evaluating data-loss-prevention and sensitive-data-detection classifiers against real school egress traffic rather than synthetic seeds; DLP policy benchmarking on which rules fire and where they over-trigger; minor-safety keyword and category corpora from chat and download activity; student data-flow compliance evidence for FERPA and COPPA audits; and shadow-AI discovery in education, an area of acute and rising procurement demand.

The data is with 5 years of history.

Coverage spans US; Application Software; alternative, sentiment, reference; equities.