Skip to content
ParamergeParamerge
Stephen Lieberman
Through Paramerge
Contact

PAN Lab example

Gaggle Safety Management

The discontinuation that changed the vendor rather than the graph: a school monitoring scanner

Gaggle's software scans students' school accounts for signs of self-harm. When students sued, one district switched vendors. Did the practice change, or only the name?

See more

Gaggle Safety Management is Gaggle.Net's monitoring service, used by roughly 1,500 US school districts. It scans students' school-issued Google and Microsoft accounts and devices around the clock, using keyword lists and machine-learning models. It flags items that might signal self-harm, violence, or other risks, and can block an email or delete a file before anyone reviews it.

How an alert is handled

Every flag is seen first by Gaggle's paid human reviewers, who work off-site and sort it, deciding which items to send on. Serious items are alerted to district safety staff: counselors, principals, and designated emergency contacts.

Content deemed to show imminent danger can trigger direct calls to the district's emergency contacts. Occasionally that ends in a police welfare check at the student's home, often late at night. The sources do not say who calls the police.

What it scans

It scans email, documents, shared drives, chats, web searches, and images. That takes in homework, college essays, art portfolios, and student newspaper files. It flags items one at a time, sorted into tiers of severity. It does not give each student a risk score.

A March 2025 investigation by the Associated Press and the Seattle Times found parents were minimally informed and opt-outs were denied. Students and families had essentially no way to appeal or opt out.

How many flags there were

In Vancouver, Washington, Gaggle raised alerts on nearly 2,200 students between October 2023 and October 2024. That is about 10 percent of enrollment, and 25 percent at one school. Counselors received three to four alerts each a month, and about half led to contact with a parent. At least six students who wrote about their sexual orientation or gender identity risked being outed.

In Lawrence, Kansas, the system logged more than 1,200 incidents in ten months. Officials deemed about two-thirds of them nonissues. That figure comes from the students' complaint, drawing on district records, and the district has not independently accounted for it.

More than 200 Lawrence alerts came from homework alone. Photography students' art-portfolio images were deleted as suspected nudity or child pornography. They were later ruled false alarms, and the students recovered them only from personal backups. A student was flagged for editing a friend's college essay that contained the words "my mental health."

Four editions of the student newspaper The Budget could not publish. Gaggle blocked flagged emails and swept confidential reporting materials from the journalists' Google Drives.

The archive and the release

Every flag writes an archived screenshot or document into Gaggle's portal. It is a lasting store of students' most sensitive disclosures.

Responding to a public-records request, Vancouver accidentally gave Associated Press and Seattle Times reporters links to nearly 3,500 unredacted flagged documents. The links had no password and no firewall. The documents held students' names beside disclosures of suicide attempts, depression, and sexual orientation.

Gaggle afterward changed its system so screenshot links require a login after 72 hours. It defends the window without a login as needed so emergency contacts "can respond quickly." That reason is Gaggle's claim.

Whether it works

A 2023 RAND Corporation review found only scant evidence of either benefits or risks from AI monitoring of this kind. Its co-author Benjamin Boudreaux observed: "If you don't have the right number of mental health counselors, issuing more alerts is not actually going to improve suicide prevention."

No data on correct flags, welfare checks, or outcomes has been published. Vancouver's summer 2024 contract was $328,036 for three school years, about the cost of one additional counselor.

There is no federal regulation specific to monitoring for school safety. Districts cite compliance with the Children's Internet Protection Act as the rationale for monitoring.

The lawsuit

Lawrence adopted Gaggle in November 2023 under a roughly $160,000 three-year contract running through July 31, 2026. In early August 2025, nine current and former students of Lawrence and Free State High Schools sued the district. They also sued its board president and its superintendent.

Most were journalists at the student papers The Budget and The Free Press, and the others were visual-arts students. They allege that the district searched and seized their speech and files without any suspicion of wrongdoing. They claim this violates the First, Fourth, and Fourteenth Amendments, and seek damages and a permanent court order.

After the filing, Lawrence High's principal briefly barred The Budget from covering the lawsuit or Gaggle, then withdrew the order. The Student Press Law Center reports that the paper's adviser was warned, through union representatives, that their job was in danger.

The switch to another vendor

Court filings revealed by late October 2025 that Lawrence had stopped using Gaggle and switched to another vendor, ManagedMethods. There was no school board vote and no public disclosure. The switch showed up only as a line in the check register, the list of payments presented at the October 27, 2025 board meeting.

The superintendent cited about $38,700 in savings a year and compliance with the Children's Internet Protection Act. The students' amended complaint, filed in November 2025, added claims under the Kansas Open Records Act. It argued that the switch does not end the case, because the core practice continues: suspicionless scanning, flagging, and seizure of student speech and belongings.

What the court has decided

On April 10, 2026, federal Judge Kathryn Vratil found the district had violated the Kansas Open Records Act. It had never substantively answered the students' late-October 2025 requests for the ManagedMethods purchasing records and the Gaggle phase-out and data-handling records. The district had claimed roughly 10,000 documents, at about $13,000 to process.

She ordered the district to respond within days, describing its search, its timeline, and any reasons for withholding. That was a deadline to respond, not an order to produce everything. She dismissed the students who had graduated, because they no longer had a legal stake in the case, and dropped two administrators as defendants.

On June 4, 2026, she ordered the district to pay the students' attorney fees for the open-records violations. She called its conduct "drawn out, hollow and perplexing" and likened it to "stonewalling." A jury trial of 8 to 10 days on the surviving constitutional claims is set for January 4, 2027. The reporting runs through July 2026, and those claims are undecided.

Other districts and vendors

Durham, North Carolina, had already dropped Gaggle in 2023, after an LGBTQ+ outing incident. GoGuardian Beacon and Securly are Gaggle's main competitors, named in the same reporting. Neither is drawn here.

What this network is drawn from

This network follows the pattern the case file describes. It does not reconstruct the actual Gaggle or ManagedMethods system, and it makes no claim about either vendor's accuracy. It shows the scanner, the students' accounts, Gaggle's reviewers, district staff, police welfare checks, the archive, the school board, the lawsuit, and the successor vendor. It also shows two checks the sources do not describe. The students themselves are outside the network.

What the available tools can and cannot address

A failure pathway is a link between two parts of the network where a mistake made by one part can be passed on to the other. A tool closes a pathway when mistakes stop passing along it. The work along it may go on.

This case has a budget of 11 units and starts with no pressure applied. Explore (No Targets) sets no targets. The other three levels all ask for the network's mistakes to be contained, meaning corrected rather than building on each other.

Under Service Targets Only, the targets can be met in hundreds of ways. The cheapest costs 3 units: Mark AI-written records alone, at its stronger setting. Many pairs cost 4 units, such as Escalate checks with Mark AI-written records or with Peer sharing rules. With lingering effects turned off in the Dynamics menu, Peer sharing rules at its stronger setting also meets them alone for 3 units.

Service and Safety Targets also asks you to close every failure pathway and keep up with the work, among other targets. Seven pathways are open before any tool is used. The first four are Scanner reads school accounts, Flags sent to Gaggle's reviewers, Alerts sent to district staff, and Serious items passed to district staff. The other three are Flag saved to the archive, Staff notes added to the flag record, and Earlier flags read at the next alert.

The targets can be met, and every way uses the same four tools, costing 9 units at the cheapest. Escalate checks closes the two links that carry flags to people. Peer sharing rules closes the hand-off from Gaggle's reviewers. Store less data closes the two links that write to the archive. Mark AI-written records closes the scanner's reading of accounts and the reading of earlier flags.

Within the budget, one more tool can join the four: Keep prompts neutral, Gate vendor updates, Review on schedule, or Keep skills sharp. That makes five tool sets, or ten ways once stronger settings are counted.

All Governance Targets asks for more again. The same four tools are needed, costing 9 units at the cheapest. Four tool sets meet it: the four alone, or with Keep prompts neutral, Gate vendor updates, or Keep skills sharp. Counting stronger settings, there are nine ways.

Vet connections is in no combination that meets the two higher levels within the budget, because adding it to the four costs 12 units. Pause AI on alarms is in none even with the budget lifted. It cuts the benefit the scanner brings too far.

More is not better here. Using every tool at its strongest setting costs 40 units, well over the budget of 11. It contains the mistakes and closes every failure pathway. But the benefit the scanner brings to the work falls below what the targets ask, so it meets them at none of the three levels.

Stylized model of a documented deploymentBehavioral-health & crisis triage

Open this example in PAN Lab v0.1 to apply pressures and levers and watch what the system does.

What this models

This example runs on the Gaggle-monitoring-class student-communication safety scanner network: 9 components and 15 pathways between them. Every context in the Lab is a stylized model, never a reconstruction of any actual deployment, and each assumption behind it carries a provenance label.

Evidence base: 3 assumed · 3 published baseline. In the Lab, the shaded evidence band behind each headline readout draws its width from the least-established class below.

Show all 6 assumptions
  • assumed

    This network follows the pattern of student monitoring documented in the Gaggle case file. It does not reconstruct the actual Gaggle or ManagedMethods system, and it makes no claim about either vendor's accuracy. It shows a scanner that flags a great deal with low precision, reading minors' whole school digital life. A chain of people handles its flags: Gaggle's reviewers, district staff, and occasionally police at a student's home. The archive of flags became the source of a data release. And the students argue that the vendor switch left the practice in place. The scanner flags items one at a time and gives no student a risk score. Some automated actions happen before any person reviews an item.

  • baseline

    The network draws three facts from the sources. First, each step in the alert chain adds its own risk. At the school step it is outing: in Vancouver, at least six students risked being outed. Durham, North Carolina, had already dropped Gaggle in 2023 after an LGBTQ+ outing incident. At the police step, the risk is a late-night visit prompted by a wrong flag. In Lawrence, officials deemed about two-thirds of more than 1,200 incidents nonissues, so staff spend much of their effort dismissing flags. That count comes from the students' complaint, drawing on district records, and the district has not independently accounted for it. Second, the archive of flags is what was exposed. Vancouver accidentally released nearly 3,500 unredacted flagged documents through links that needed no password. The network also draws the window in which alert links open without a login. Third, two checks the sources do not describe are drawn as checks. One is an independent evaluation of accuracy and of whether the monitoring works. The other is a standing school board review with a way for students to appeal.

  • baseline

    The vendor switch is the case's defining event. The network draws it as the successor vendor outside the district, and as a school board approval the switch went around. Lawrence stopped using Gaggle during the lawsuit and switched to another vendor, with no school board vote and no public disclosure. The switch showed up only as a line in a check register, the list of payments presented at a board meeting. The superintendent cited about $38,700 a year in savings and compliance with the Children's Internet Protection Act. The students argued that the switch does not end the case, because suspicionless scanning, flagging, and seizure of student speech and belongings continue. The network carries that as the students' position and as the documented sequence of events. It does not decide the constitutional claims, which go to a jury trial set for January 2027.

  • baseline

    The network draws the lawsuit and the open-records rulings as a check from outside the district. The school board and its purchasing process were bypassed. So the checks that forced a response came from outside: a federal lawsuit and two rulings under the Kansas Open Records Act. On April 10, 2026, a federal judge found the district had violated that law. It had never substantively answered the students' requests for records on the switch and on Gaggle's phase-out. On June 4, 2026, she ordered the district to pay attorney fees, calling its conduct "drawn out, hollow and perplexing." The scanner had swept the reporting files of the student journalists who became plaintiffs. After they sued, the principal briefly barred their paper from covering the lawsuit, and later withdrew the order. The Student Press Law Center reported it. Whether the monitoring works is unestablished either way: a 2023 RAND review found only scant evidence. So the network claims no benefit or harm from it.

  • assumed

    The network marks private data on six kinds of link. They are the scanner reading accounts, reading and writing the archive, and the escalation to police. The others are documents leaving the district, records sought under open-records law, and the window without a login. The monitoring covers minors' whole school digital life around the clock. That includes email, documents, drives, chats, web activity, images, homework, college essays, art portfolios, and journalism. Families had little or no way to opt out. The scanner flags disclosures of sexual orientation and mental health. The Vancouver release exposed nearly 3,500 unredacted flagged documents at once. Gaggle afterward required a login on screenshot links after 72 hours. Its reason for the window without a login, that emergency contacts can respond quickly, is Gaggle's claim. The tools that act on these links are Store less data, Vet connections, and Mark AI-written records.

  • assumed

    Students are outside the network. No clinical, disciplinary, outing, or welfare-check outcome for any student is computed. A flag, an alert, or a welfare check here stands for an institutional signal, never a person. The two-thirds figure from Lawrence records what officials decided about flags afterward. It comes from the students' complaint and district records, and it is not a measured error rate. No figures have been published on how often the scanner or the reviewers are right, how many welfare checks happen, or what outcomes follow. RAND found only scant evidence either way. Vancouver's figures, the $328,036 contract and the roughly 2,200 flagged students, are kept apart from Lawrence's roughly $160,000 contract. GoGuardian Beacon and Securly, competing products named in the same reporting, are context only and are not drawn here.

What this example does not show

Show all 5 limitations
  • This example does not show any student's outcome. It shows how mistakes pass between the scanner, the people who handle its flags, and its records, and students stay outside it. A flag, an alert, or a welfare check here is an institutional signal, never a person. The real harms fall on students: the risk of being outed, late-night police welfare checks at home, and a student newspaper unable to publish. They are recorded in the case file and never computed here.
  • This example does not show whether the monitoring works. A 2023 RAND review found only scant evidence of either benefits or risks. Gaggle publishes no figures on how often its flags or its reviewers are right, and no welfare-check counts exist. The two-thirds figure from Lawrence comes from the students' complaint, drawing on district records, and the district has not independently accounted for it. It is not a measured error rate. How this network looks when it opens is not a finding about any real deployment's accuracy.
  • This example does not decide the constitutional claims. They are unresolved, and a jury trial is set for January 2027. The vendor switch and the open-records rulings are reported as the documented sequence of events. The students' argument, that switching vendors leaves the scanning practice in place, is their position. The district's reasons, cost savings and compliance with the Children's Internet Protection Act, are reported beside it. Gaggle's own claims are labeled as its claims. They include its chief executive's framing of monitored accounts as a digital playground without fences. They also include its defense of the window without a login, and its account of the change it made after the release.
  • This example keeps each district's figures to that district and date. The Vancouver release of nearly 3,500 unredacted documents, the roughly 2,200 flagged students, and the $328,036 three-year contract belong to Vancouver Public Schools in Washington. The 1,200-plus incidents, the roughly $160,000 contract, and the vendor switch belong to the Lawrence school district, USD 497, in Kansas. The two are not combined. GoGuardian Beacon and Securly, competing products with the same chain from scanner to school to welfare check, are named in the same reporting. They are context only, not drawn here.
  • This example states the outing risk as exposure only: what the archive holds and which hand-offs can pass it on. No outing, disciplinary, welfare-check, or family outcome for any student is shown or computed. The 2023 Durham discontinuation after an outing incident, and the students who risked being outed, come from the cited reporting, not from this network. The general point, that a store built to protect people can become how they are exposed, comes from a 2026 academic book chapter on AI and LGBTQIA+ people. The chapter also records practitioners leaving sexual-orientation and gender-identity data out of records on purpose, at a cost to completeness. It gives a direction, with no size attached.

Sources and evidence

What this example rests on, claim by claim. Every entry resolves to the same ledger the Evidence Registry publishes.

  • Gaggle's student-communication safety monitoring, used by roughly 1,500 US districts covering about 6 million students as of a March 2025 AP and Seattle Times investigation, scans school-issued accounts around the clock and routes flags through a multi-hop chain (a machine flag, an off-site vendor reviewer, district safety staff, and, for imminent-danger after-hours alerts, occasional police welfare checks); in Vancouver Public Schools nearly 2,200 students (about 10% of enrollment) triggered alerts in one year, in Lawrence USD 497 more than 1,200 incidents were logged in ten months with about two-thirds deemed nonissues by officials (a figure the plaintiffs drew from district records), and Vancouver accidentally released the archive of flagged documents to reporters as nearly 3,500 unredacted files through unprotected links, while a 2023 RAND review found only scant evidence of either benefit or risk and the vendor publishes no accuracy figures.

    empirical
    • Investigative Bryan and Lurye, Schools use AI to monitor kids. An investigation found security risks (The Christian Science Monitor, AP and Seattle Times Education Reporting Collaborative, 2025) https://www.csmonitor.com/USA/Education/2025/0312/ai-surveillance-schools-gaggle
    • Investigative Associated Press, Takeaways from our investigation on AI-powered school surveillance (2025) https://srnnews.com/takeaways-from-our-investigation-on-ai-powered-school-surveillance/
    • news Lawrence Journal-World, Lawrence school district sued in federal court for use of AI-powered surveillance system; students claim Gaggle results in illegal searches (2025) https://www2.ljworld.com/news/schools/2025/aug/07/lawrence-school-district-sued-in-federal-court-for-use-of-ai-powered-surveillance-system-students-claim-gaggle-results-in-illegal-searches/

  • After nine Lawrence, Kansas students sued their district in early August 2025 over its use of AI communication monitoring, court filings revealed the district had ceased using Gaggle mid-litigation and substituted a different monitoring vendor with no board vote or public disclosure — surfacing only as a line in a check register — and the plaintiffs' amended complaint argued the swap does not moot the case because the core practice of suspicionless scanning, flagging, and seizure of student speech continues; on April 10, 2026 a federal judge found the district violated the Kansas Open Records Act in withholding the substitution and phase-out records, and on June 4, 2026 ordered it to pay the students' attorney fees, characterizing the conduct as drawn out, hollow, and perplexing, with a jury trial on the surviving constitutional claims set for January 2027.

    empirical
    • news Heimsoth, Students allege continued unconstitutional AI digital surveillance with new vendor and violations of Open Records Act in school district lawsuit (Lawrence Journal-World, 2025) https://www2.ljworld.com/news/schools/2025/nov/21/students-allege-continued-unconstitutional-ai-digital-surveillance-with-new-vendor-and-violations-of-open-records-act-in-school-district-lawsuit/
    • news Heimsoth, Federal judge finds Lawrence school district violated open records law in student lawsuit regarding Gaggle (Lawrence Journal-World, 2026) https://www2.ljworld.com/news/schools/2026/apr/10/federal-judge-finds-lawrence-school-district-violated-open-records-law-in-student-lawsuit-regarding-gaggle/
    • news Heimsoth, Federal judge orders Lawrence school district to pay attorney fees to students in Gaggle case after KORA violations (Lawrence Journal-World, 2026) https://www2.ljworld.com/news/schools/2026/jun/04/federal-judge-orders-lawrence-school-district-to-pay-attorney-fees-to-students-in-gaggle-case-after-kora-violations/

Where this connects

Institutional pressures in this domain

  • Workload surge — Demand outruns staffing; per-case attention shrinks and review becomes triage.
  • Reviewer bottleneck — One fixed-capacity checking stage sits between AI output and consequence; everything queues behind it.
  • Vendor opacity — The deploying institution cannot inspect the model, data, or update pipeline it is accountable for.
  • Deadline pressure — Statutory or managerial timeliness rules reward fast approval of machine output over slow disagreement.
  • Compliance over substance — Paper controls (sign-offs, checklists) satisfy audits while the behavior they describe erodes.

All of them in context on the Behavioral-health & crisis triage domain page.

Levers available here and the patterns behind them

Documented case histories