evidoria

← Back to browse

Good practice Imported

Puerto Rico's DEPR Scales Microsoft Copilot to 220,000+ Students — AI-Assisted IEPs Support Students with Disabilities

Puerto Rico · San Juan · See the Puerto Rico profile · See the San Juan profile

Evidence: Descriptive / self-reported Top 25% 60/100 · Ask Evidence Copilot about this practice

Puerto Rico's Department of Education rolled out Microsoft Copilot to 20,000+ teachers and 220,000+ students, cutting assessment-writing time from an hour to 15 minutes and using AI to draft individualized education plans and dual assessments for neurodivergent students.

220,000+ students
Students in system (2025-2026)
20,000+ people
Teachers/administrators (2025-2026)
15,000 licences
Active Copilot licences issued (2025-2026)
200,000+ devices
Windows devices in system (2025-2026)
40,000 devices
iPads in system (2025-2026)
~60 to ~15 minutes
Assessment-writing time reduction (2025-2026)

Details

Maturity
Scaling
Promoter
Departamento de Educación de Puerto Rico (DEPR) with Microsoft
Period
2025–2026
Keywords
generative AI, special education, teacher productivity, inclusion

Context

Puerto Rico's public school system serves more than 220,000 students. Its Department of Education (DEPR) set out to give overstretched teachers, including special-education staff, a practical AI assistant rather than impose a top-down mandate.

Objectives

DEPR partnered with Microsoft to roll out Microsoft 365 Copilot starting in 2025, aiming to reduce teacher administrative burden and support inclusion — including drafting individualized education plans (IEPs) and dual assessment versions for neurodivergent students.

Activities

Building on an existing base of over 200,000 Windows devices and 40,000 iPads, DEPR issued 15,000 active Copilot licences to its more than 20,000 teachers and administrators. The rollout is paired with an age-gated ethical-use framework ('Modelo Sense') restricting AI use for younger children, and an initial cohort of 15 schools combines AI with esports, robotics and CertiPort certification.

Results

Teachers report that assessment and evaluation creation time fell from roughly one hour to about 15 minutes using Copilot.

Conclusions

The reported figures — licence counts, device totals and time savings — come from department and vendor accounts rather than an independent evaluation, and no rigorous study yet measures student learning outcomes or whether AI-assisted IEPs meet special-education quality standards; this should be read as an early, large-scale implementation case rather than a proven-impact study.

Implementation

Indicative cost
High (€500k–€5M)
Time to results
Short (< 1 year)

Commonly funded by

National / regional programmes

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