evidoria

← Back to browse

Good practice Imported

Alongside AI Chatbot Fills Counselor Gaps in US Schools — Northwestern Pilot Finds Mixed Results

United States of America · Corsicana · See the United States of America profile

Evidence: Observational / pre–post Top 89% 27/100 · Ask Evidence Copilot about this practice

Alongside, an AI chatbot now in 200+ US schools across 19 states, helped flag at-risk students in a Texas/New Mexico pilot, but showed no significant effect on depression, anxiety or loneliness overall, and an independent Common Sense Media review found real detection gaps.

200+
Schools using Alongside (2026)
19
US states covered (2026)
~31,000
Students using Alongside (past year)
28,000
Combined usage hours (past year)
1,800+
Students flagged for immediate intervention
76 %
High-risk students reporting no suicidal ideation after 3 months

Details

Maturity
Scaling
Promoter
Alongside, Inc.
Period
2022-2026
Keywords
school counseling, student mental health, AI chatbot, K-12 administration

Context

Amid a nationwide shortage of school counselors, Seattle-based company Alongside (founded 2022) built an AI chatbot, named 'Kiwi' to students, designed to give K-12 students always-available emotional support and to flag mental-health crises to human staff.

Activities

By 2026 Alongside operated in more than 200 schools across 19 US states under a per-student subscription model. Documented sites include Corsicana High School near Dallas-Fort Worth, Texas (in use 3 years, about 1,800 students, 3 behavioural support counsellors) and Interlachen Jr-Sr High School in Putnam County, Florida (400+ students in grades 7-8).

Results

In the past year, nearly 31,000 students used Alongside for a combined 28,000 hours; the platform flagged more than 1,800 students as needing immediate intervention, and a clinical team separately reviewed 6,000 additional borderline chats that had not triggered an alert. A Northwestern University pilot study of Alongside in Texas and New Mexico schools found that 76% of high-risk students reported no suicidal ideation after three months of use, but across the full sample there was no significant effect on depression, anxiety or loneliness.

Conclusions

An independent risk assessment by Common Sense Media's Youth AI Safety Institute, examining the same pilot, documented test scenarios in which Alongside failed to detect signs of suicide risk, misread context (for example, confusing 'shooting humans' with 'shooting hoops'), and struggled to identify signs of psychosis. The case illustrates both the appeal and the limits of AI-based crisis triage in under-resourced schools: real capacity relief and some flagged interventions, but no rigorous evidence yet of improved mental-health outcomes at scale, and documented failure modes in exactly the safety-critical function the tool is meant to serve.

Implementation

Indicative cost
Medium (€50k–€500k) — Per-student subscription model; specific pricing not disclosed.
Time to results
Medium (1–3 years) — Founded 2022; documented sites in use up to 3 years; pilot study covered a 3-month usage window.
Staffing & skills
Alongside (company), school behavioural-support counsellors

Conditions for success

  • Always-available (24/7) design to fill counselor-shortage gaps
  • Clinical team reviewing borderline chats not caught by automated alerts

Common failure modes

  • No significant effect on depression, anxiety or loneliness across the full pilot sample
  • Independent risk assessment found the tool failed to detect signs of suicide risk in test scenarios, misread context, and struggled to identify signs of psychosis

Where it fits

Governance type
private company, school-district subscription model
Scale
200+ schools across 19 US states
Income level
high-income

Commonly funded by

Own resources / municipal budget

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful