DCPS Technology Infrastructure and AI Policy
Summary
DC Public Schools faces two related challenges: a persistent gap between agreed-upon technology priorities and actual implementation, and the absence of a substantive policy framework for AI use in classrooms. These are not separate problems. The organizational capacity that has delayed basic system integration for years is the same capacity that will need to respond to AI — a much faster-moving issue with more immediate consequences for teaching and academic integrity.
Why this matters
A two-year timeline to implement a single grade-passback integration, on a request both sides agreed was a priority from the outset, illustrates a structural limitation in execution — not a lack of awareness of the problem. This same limitation will directly affect how, and whether, DCPS can respond to AI in schools, where the pace of change is measured in months, not years.
Representative Examples
Grade Passback (Canvas to Aspen)
Identified as a shared priority in March 2024. Testing was first promised for spring 2024, then postponed to fall 2024, then further postponed with no full rollout expected before January 2025. As of the most recent update, it remains unimplemented.
Attendance Excuse Forms
Jackson-Reed presented data showing that fewer than 3% of actual excuse notes were being successfully recorded in Aspen, due to reliance on paper submissions. This was acknowledged as a priority for the 2025–26 school year — roughly two years after the issue was first raised.
Course Registration
No centralized online course catalog existed until Jackson-Reed's own student-run Website Club built one independently. DCPS is only now, in 2025–26, working toward a fully online registration process districtwide, and Aspen is still not being used for master schedule planning even where it is the system of record.
Attendance Tracking
Each school currently uses its own intake method, with data manually entered into Aspen and additional "shadow systems" — school websites, paper records, and phone calls — still handling work the central system should support.
Observed at JR
- At a demonstration class visit to Jackson-Reed's IT Academy, only one student in the room had used AI in any structured, instructional context.
- IT Academy enrollment has declined significantly — notable because this is the student population that should be most engaged with AI as a career-relevant skill.
- In the absence of structured instruction, AI use among students is largely happening unsupervised, most often in take-home assignments, where it does the most damage to actual learning.
A note on approach
The smartphone restriction worked because it was a subtraction — removing a distraction with a clear, enforceable boundary. AI does not lend itself to the same approach, since students have access to it outside of school regardless of any in-school policy. A prohibition-only approach risks pushing use further out of view rather than addressing it. The more workable model is sequencing: students build foundational skills first, then receive deliberate instruction in AI as a tool, with the extent of integration increasing through high school as AI fluency becomes more directly relevant to college and career readiness. A middle school policy and a high school policy should not look the same, and currently no distinction exists at all.
Two Distinct Problems Requiring Two DIfferent Responses
- Instruction — teaching students to use AI thoughtfully and effectively once foundational skills are in place. This is a curriculum and pedagogy question, addressable at the classroom and school level.
- Integrity — verifying that submitted work reflects a student's own understanding, in an environment where AI-generated work is often indistinguishable from a student's own. This is an assessment design and policy question, and the more urgent of the two, since the current absence of policy is actively rewarding the behavior it should be discouraging.
Closing Note
DCPS's technology infrastructure challenges and its lack of AI policy are not separate problems — they are the same organizational gap surfacing in two different places, one with years to catch up and one that cannot wait that long. Addressing AI policy effectively will likely require the district to first confront why implementation of far simpler, already-agreed-upon initiatives has taken years rather than months.