Table of Contents
DC Stoplight AI Policy for Schools: What's Red, Yellow, Green
The DC stoplight AI policy for schools bars AI from grading, discipline, IEP eligibility and teacher evaluations. Here is the full red-yellow-green list.
Most school AI policies published in 2026 governed what students may do. Washington, DC published one that governs what adults may do to students. The DC stoplight AI policy for schools, released September 1, 2026 by the Office of the State Superintendent of Education, sorts staff AI uses into three buckets: red for never, yellow for limited use with safeguards, green for permitted with human review. Deciding IEP eligibility is red. Teacher performance evaluations are red. Grading student work is yellow. It is guidance, not a mandate, and it reads like the most useful single page any parent could hand to a principal.
Key Takeaways
- OSSE released the AI Model Policy for Staff Use on September 1, 2026, for the 2026-27 school year. It is non-mandatory; local education agencies can customize and adopt it as they see fit.
- Red (prohibited): physical surveillance of students and staff, student discipline decisions, teacher performance evaluations, and IEP eligibility or Section 504 accommodation determinations.
- Yellow (limited, with safeguards): monitoring digital activity on agency-issued devices, drafting IEP language, reviewing and grading student work, and supplemental educator coaching.
- Green (permitted with human review): drafting lesson plans, customizing student-facing materials, developing tutoring plans, analyzing data sets, preparing communications, and supporting logistics.
- The policy covers staff use, not student use or tool procurement. OSSE’s February 2026 survey found only 45% of DC local education agencies had a staff AI policy at all.
What the DC stoplight AI policy for schools actually contains
A model policy is a template a state agency writes so local agencies don’t have to start from a blank page. DC’s covers staff use, which is a deliberate and unusual scope choice.
OSSE’s announcement frames the stoplight as clarifying where AI should not be used, where limited use may be appropriate with enhanced oversight, and where use is permitted with professional awareness and human review. Deputy Mayor Paul Kihn: “AI should enhance teaching and operations, but it should never be used to replace professional judgment.” Dr. Antoinette S. Mitchell added that the policy “keeps human judgment at the center of decision-making, strengthens privacy and data protections.”
WTOP’s coverage noted the red-light logic: any purpose that could compromise safety or privacy, including determining an individualized education program for students with disabilities, teacher performance evaluations, or physical surveillance.
The context is the gap. OSSE’s February 2026 survey of local education agency leaders found only 45% had established a staff AI policy. This document exists because most schools had nothing written down.
The stoplight table
| Use | Color | Why it lands there |
|---|---|---|
| Physical surveillance of students and staff | Red | Safety and privacy risk; no legitimate instructional purpose |
| Student discipline decisions | Red | Consequences require human judgment and due process |
| Teacher performance evaluations | Red | High-stakes employment decision about a person |
| IEP eligibility or Section 504 accommodation determinations | Red | Legally consequential determination about a child’s rights |
| Monitoring digital activity on agency-issued devices | Yellow | Permissible with safeguards; scope and notice matter |
| Drafting IEP language | Yellow | Drafting is different from deciding; a human must own the content |
| Reviewing and grading student work | Yellow | Allowed with oversight; the teacher remains responsible for the grade |
| Supplemental educator coaching | Yellow | Useful as a supplement, not a substitute for a mentor |
| Drafting lesson plans | Green | Teacher reviews and adapts before use |
| Customizing student-facing materials | Green | Human review required before it reaches a child |
| Developing tutoring plans | Green | Plan, not placement; a human executes it |
| Analyzing data sets | Green | Analysis informs a decision rather than making one |
| Preparing school communications | Green | Routine drafting with review |
| Supporting logistical operations | Green | Scheduling, routing, and similar back-office work |
The distinction that does the most work is between drafting and deciding. Drafting IEP language is yellow; determining IEP eligibility is red. That line is the single most portable idea in the document, and it generalizes: a machine can produce a sentence a human then owns, but it should not produce a conclusion about a child that a human merely signs.
Why “who decides” matters more than “which tools”
Most parent anxiety about school AI is about students using chatbots. The policy category that will actually affect your child’s transcript is AI used on them.
The clearest evidence for caution comes from the detector record. In May 2026, a Wake County, North Carolina freshman received a zero on an English assignment after her teacher ran it through three AI detectors; she appealed, a second teacher reviewed her work, and the grade was changed to 100. By June, the district’s draft policy stated it does not support AI detection programs “due to their technical unreliability, inaccuracy, and potential for bias” against English learners. A month earlier, a Purdue professor accused more than 200 students of AI use and withdrew the allegations within days. In both cases, a tool produced a conclusion and a human treated it as a finding. That is precisely the failure mode DC’s red category is written to prevent.
Other jurisdictions arrived at similar lines by different routes. Oklahoma’s SB 1734 bars AI from being “primarily used for grading, discipline or other high-stakes educational decisions” and gives parents an opt-out with no academic penalty. New York City’s September 2026 policy permits teacher AI use for lesson planning, translation, and family communications while prohibiting it for grading, behavior monitoring, counseling, and special-education planning. Charleston County, South Carolina’s board-approved policy from January 2026 prohibits AI as the sole basis for high-stakes decisions including discipline, placement, and special education determinations.
Four separate bodies drew nearly the same line in the same year. When that happens, the line is probably real.
The honest limits: DC’s policy is non-mandatory, it addresses staff use rather than student use or procurement, and no one has outcome data on whether stoplight frameworks change behavior. A model policy that a local agency declines to adopt changes nothing.
What to do with the stoplight at your own school
Hand it over as a list, not an argument
The most effective version of this conversation is not “I read about AI and I’m worried.” It’s “Washington, DC’s education agency published a model policy that bars AI from grading, discipline, teacher evaluations, and IEP eligibility decisions. Does our district’s policy address those four?” A specific list invites a specific answer.
Ask the IEP question directly if your child has a plan
The red/yellow split on IEPs is the most consequential row for families in special education. Ask, in writing: is AI used anywhere in eligibility determination, and if AI is used to draft IEP language, who reviews and owns the final text? Put the answer in the IEP record. That is a documentable service question, not a philosophical one.
Ask what happens when a tool flags your child
Yellow-light grading and digital monitoring both produce flags. Ask what the process is after a flag: who reviews it, how a parent is notified, and how a student contests it. The Wake County case resolved well because an appeal existed and a second human looked. Find out whether yours does.
Ask about physical surveillance specifically
DC put physical surveillance in red. Many districts have cameras with analytics, and some have facial recognition. Ask whether any camera system uses AI analysis and, if so, what it flags and how long footage is retained. Our ed-tech app privacy audit covers the data questions; this one is worth asking separately because it rarely appears in an app inventory.
Use the drafting/deciding line at home too
The distinction is a good household rule with a different label. Your kid can ask AI to draft a sentence they then own; they cannot ask it to decide what they think. That mapping makes school policy legible to a 12-year-old, and it happens to be the same principle the adults are being held to.
What not to do
Don’t assume a policy that governs student use also governs staff use. They’re separate documents in most districts, and the staff one is usually newer and thinner. And don’t lead with the assumption that your school is misusing AI. OSSE’s 45% figure suggests the more common situation is that nobody has written anything down yet, which is a solvable problem and a much easier conversation.
What to Watch For Over the Next 3 Months
- Week 4: Ask whether your district has a staff AI policy and whether it addresses the four red-light uses. If your child has an IEP or 504 plan, ask that question in writing and keep the reply.
- Month 2 red flags: Grade changes or placement changes with no named human decision-maker. Feedback on assignments that reads as machine-generated with no teacher comment. A district that can’t say whether its camera systems use AI analysis. An academic-integrity process that starts and ends with a detector score.
- Month 3 self-check: Can you name, for your child’s school, who decides a discipline case, who decides IEP eligibility, and whether any software is involved in either? Those three answers are the whole point of a stoplight framework. Watch too for whether DC local agencies actually adopt the model, since it is voluntary.
Frequently Asked Questions
Is DC’s stoplight policy binding on schools?
No. It is a model policy that local education agencies may customize and adopt as they see fit, released for the 2026-27 school year. Its value is as a template and a benchmark you can ask your own district to match.
Does it cover what students are allowed to do with AI?
No. The scope is staff use. It does not address student use or tool procurement, which are separate policy areas. For student-facing rules, Katy ISD’s grade-by-grade framework and New York City’s moratorium are the clearer examples.
Why is grading yellow rather than red?
Because the policy distinguishes between AI assisting a review and AI making a determination. Reviewing and grading student work sits in the limited-use category with enhanced oversight, meaning the teacher remains responsible for the grade. Discipline and IEP eligibility, where the output is itself a consequential decision, are red.
What does the IEP distinction mean in practice?
Drafting IEP language is yellow: a staff member may use AI to help write text they then review and own. Determining IEP eligibility or Section 504 accommodations is red: that decision must be made by people. If your child has a plan, that’s the line to confirm in writing.
Do other places draw the same lines?
Yes, with variations. Oklahoma’s SB 1734 bars AI as the primary basis for grading, discipline, or other high-stakes decisions. New York City prohibits teacher AI use for grading, behavior monitoring, counseling, and special-education planning. Charleston County, South Carolina prohibits AI as the sole basis for discipline, placement, and special education determinations.
My district uses an AI detector on student essays. Where does that fall?
Detection is a review tool whose output can become a disciplinary finding, which puts it uncomfortably close to the red category. The 2026 record is not encouraging: a Wake County student got a zero from three detectors and a 100 on appeal, and her district subsequently rejected detector use. Ask what corroborating evidence is required before a detector score affects a grade. We cover the rights angle in AI cheating detectors are failing students.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- Office of the State Superintendent of Education, District of Columbia. (2026, September 1). “OSSE Releases AI Model Policy to Guide Responsible Staff Use in Schools.” https://osse.dc.gov/release/osse-releases-ai-model-policy-guide-responsible-staff-use-schools
- WTOP News. (2026, September). “No, robots won’t be teaching your kids. DC releases new AI guidelines for teachers.” https://wtop.com/dc/2026/09/no-robots-wont-be-teaching-your-kids-dc-releases-new-ai-guidelines-for-teachers/
- K-12 Dive. (2026, July 9). “4 more states require districts to adopt AI policies.” https://www.k12dive.com/news/4-more-states-require-districts-to-adopt-ai-policies/824749/
- Office of the Mayor of New York City. (2026, September 2). “Nation’s Broadest Generative AI Moratorium in Schools.” https://www.nyc.gov/mayors-office/news/2026/09/mayor-mamdani-and-chancellor-samuels-put-students-first-with-nat
- WRAL. (2026, May). “Wake County student says clear AI policies needed after being accused of cheating.” https://www.wral.com/news/education/wake-county-student-says-ai-policies-needed-after-cheating-accusation-may-2026/
- Bailey, J. (2026, April 22). “Cheating Allegations Lead to Chaos at Purdue University.” Plagiarism Today. https://www.plagiarismtoday.com/2026/04/22/cheating-allegations-lead-to-chaos-at-purdue-university/
- Common Sense Media. (2026, August 18). Teens in the AI Era: Schoolwork and the Skills That Matter. https://www.commonsensemedia.org/research/teens-in-the-ai-era-schoolwork-and-skills-that-matter
- WRAL. (2026, June). “No AI detectors, more citations. What’s in a new Wake schools’ AI policy draft.” https://www.wral.com/news/education/whats-in-wake-schools-new-ai-policy-draft-june-2026/
- Pursuit. (2026). “Latest AI in Education News and Policies.” https://www.pursuit.us/news/ai-in-education-news-policies-innovations
