SchoolAI screens student conversations for safety concerns and flags the ones worth a closer look. Aggregate Alert Reporting gives district and school leaders a single view of alert activity across their organization to streamline visibility and prioritize student safety.
What is Aggregate Alert Reporting?
The “Alerts” tab on your Usage Report shows how many student messages were flagged during a period you choose, set against how many chat messages were screened. You can break alerts down by school, source, educator, or student, browse the individual alerts, and click any alert to open the full conversation in Chat Audit.
A few things to know
It shares filters, saved views, and export tools with the rest of the Usage Report, so you may already be familiar with how to use it.
Only Owners and Admins can see the Usage Report, including the Alerts tab.
Alert emails haven't changed. They still go to Space collaborators, and now they deep-link into this view.
Where to find it: Click your avatar in the bottom-left corner of SchoolAI, select “Usage report,” then open the Alerts tab.
How to read alert volume
An alert means a conversation was flagged for review, it does not mean something bad happened.
SchoolAI's safety screening is deliberately built to over-report rather than under-report. Missing a real concern is far worse than flagging a conversation that turns out to be fine.
Some alerts will turn out to be safe on a closer look. That is the system working as designed, not a defect.
Alert volume is not incident volume. More usage produces more alerts. A district with more alerts is not necessarily a district with more problems.
Always read the percentage. The Total alerts card shows the count alongside the share of chat messages that had an alert, so you can see the proportion rather than just the number.
Treat an alert as a prompt to look, not a conclusion. Review it, then decide.
When you share this view with a board or leadership team, set that framing first. "Here's how many conversations we screened, here's how many we flagged for a closer look, and here's what we found" is a very different conversation from "here's how many incidents we had."
How it works
1. Choose your scope
Use the search bar and filters at the top of the Alerts tab to narrow what you're looking at. Everything on the tab updates as you apply them.
Search alert text — find a specific alert by what it says
Date range — a school year or a custom range
Alert type
School
Source — Sidekick and/or specific Space names
Educator — the owner of the session
Student
Tip: If you are ever unsure what you are currently looking at, check which filters are applied. Clear your filters to get back to the full organization view.
2. Read the Total alerts card
The Overview section summarizes the period you selected with six cards:
Card | What it counts |
Total alerts | The number of alerts, with the percentage of chat messages that had an alert (for example, “1.1% of 75,733 chat messages”) |
New in the last 7 days | Recent alerts, compared with the week before |
Safety and Violence | Alerts in those two categories combined |
The percentage is the number to anchor on. It turns a raw count into a proportion.
3. Break down the alert rate
The Alert rate by school card shows the share of student messages with an alert. Use the toggle at the top of the card to change the breakdown:
School — compare buildings
Source — see whether alerts are coming from Sidekick or from specific Spaces
Educator — see alerts by session owner
Student — see alerts by student
Each row shows the number of alerts and the alert rate, with a color-coded bar by alert type. The legend above the bars shows the total for each type. Select a school or alert type to filter the rest of the tab to it.
Alert types are Safety, Violence, Inappropriate content, and Other. Select Trend to see how alert volume changes over time.
Note: Every alert since April 16, 2026 uses these four standard categories. If your date range goes back before that date, you may see older, AI-generated categories for earlier alerts.
4. Browse the alerts behind the numbers
Below the summary, the alerts table lists each individual alert, most recent first. For each one you can see:
Detected — when the alert happened
Student — who was involved
Alert type — Safety, Violence, Inappropriate content, or Other
Alert — a summary of what was flagged and why
School
The table shows 25 alerts at a time and tells you how many match your current filters. Use the filters and search to narrow it down.
Looking for a pattern at one building? Filter by school.
Want to know which Space is generating alerts? Filter by source.
Following up with a teacher about their class? Filter by educator.
Checking in on a specific student? Filter by student.
5. Review an individual conversation
Click any alert to open it in Chat Audit. At the top you'll see the student, the educator, and links to the Space and the session. Switch between the Chat, Alerts, and Insights tabs to read the conversation and see what was flagged. The Alerts tab shows the alert type, when it was detected, and why it was flagged.
Reading an alert in the context of the messages around it usually changes the interpretation, which is exactly why the review step exists. Select Back to alerts to return to your filtered list.
6. Save the views you review regularly
If you do a weekly or monthly alert review, save your filter configuration as a named view so you can reopen it in a click. Saved views on the Alerts tab work the same way as on the Usage tab, including sharing a view with other admins.
7. Export for reporting
Both export paths let you choose which tabs to include and default to Alerts when you're on the Alerts tab:
Print — opens a print dialog you can also use to save as a PDF
CSV — downloads the underlying data. If you select both Usage and Alerts, you'll get two separate files
You can also open a chat against the report to ask questions in plain language or change your filters conversationally.
What Aggregate Alert Reporting can help you do
Answer "are our students safe on this platform" in a board meeting with volume, proportion, and trend rather than anecdote
Put a single alert email in proportion when a board member or parent asks about it
See whether a school has a recurring pattern worth addressing at a staff meeting
Trace a cluster of alerts back to a specific Space, educator, or source
Document safety monitoring for a district safety review or a compliance conversation
Establish a monthly alert review cadence with a saved view and a CSV export
Frequently asked questions
Why am I seeing so many alerts? Alert volume reflects how many messages were flagged for review, not how many incidents occurred. The system errs toward flagging. Read the count against the percentage of chat messages shown on the Total alerts card.
Does an alert mean something bad happened? No. An alert marks a conversation worth reviewing. Some will turn out to be safe on a closer look — that's expected.
Why don't I see alerts from earlier in the year? Alert data in this view is available starting July 1, 2026. Conversations before that date aren't represented here.
Can I see the conversation behind an alert? Yes. Click any alert to open the full conversation in Chat Audit, with links to the Space and session.
Can I see which student an alert came from? Yes. The alerts table shows who was involved, and you can break down or filter alerts by student.
Who receives alert emails? Email behavior is unchanged. Alerts continue to go to Space collaborators and emails now deep-link into this view.
Who in my district can see the Alerts tab? Owners and Admins, the same as the rest of the Usage Report.
Can I take action on an alert from this view? You can click through to Chat Audit to review the conversation. Actions on a specific conversation — following up with a student, adjusting a Space — happen from there.
Can I dismiss an alert? Yes. Open the alert in Chat Audit and select Dismiss. If a flagged conversation was clearly fine, you can also use the in-app Feedback button to tell us — that signal helps improve screening over time.
Do filters apply to everything? Yes. Filters update the entire tab so the story stays consistent.
Why do some older alerts show different categories? Alerts from before April 16, 2026 may show older, AI-generated categories. Since then, every alert uses four standard categories: Safety, Inappropriate content, Violence, and Other.
Are alerts generated for Tools, Assistants, or Dot outside of Spaces and Sidekick? Not in this view. Aggregate Alert Reporting currently covers Spaces and Sidekick conversations.


