Chicago Public Schools Gemini Chatbot Rollout: AI Policy

Chicago Public Schools Gemini Chatbot Rollout: AI Policy

The two CPS policy resources reviewed for this report do not mention Google Gemini or a named chatbot vendor, proposed, piloted, or canceled. Claims about a Chicago Public Schools Gemini chatbot rollout have circulated, but nothing in the district's published AI guidance or its student use policy backs that up.

What CPS has published instead is a governance framework for artificial intelligence in its schools, one the district says it has worked on since the 2024-25 school year and reviews every three months. That framework, not any vendor claim, is the part of this story a district, teacher, or parent can actually verify and put to use.

What this review checked, and what it didn't

This report covers two CPS resources: the district's "AI at CPS" page and Board Report 25-1023-PO2, the Student Acceptable Use policy adopted Oct. 23, 2025. It did not include a search of procurement records, contract databases, or board meeting agendas beyond these two published policy pages, and it did not include direct outreach to CPS or Google for comment.

That means the finding here is narrow. A Gemini reference doesn't appear in CPS's public AI guidance or its student use policy as of this date. That's different from confirming no internal pilot, vendor conversation, or unpublished decision ever took place.

Anyone trying to confirm or deny a specific chatbot deployment at CPS should contact the district's Department of Information and Technology Services or Office of Teaching and Learning directly, since those two offices jointly issue the district's AI guidance (CPS, "AI at CPS"). A parent or reporter chasing this story further would also want CPS's procurement portal and recent board meeting minutes, neither of which this review searched.

What CPS policy says about a Chicago Public Schools AI chatbot

CPS's generative AI guidance governs employees, students, contractors, volunteers, and third-party vendors across the district (CPS, "AI at CPS"). The district says it has worked to implement and refine that guidance since the 2024-25 school year (CPS, "AI at CPS").

CPS also says it will review the guidance every three months to account for legal changes, technology shifts, and feedback from school communities (CPS, "AI at CPS"). Because of that schedule, anyone citing CPS's AI rules should check the current version rather than rely on an older policy summary, including this one.

The Student Acceptable Use policy, adopted Oct. 23, 2025, added artificial intelligence and media-literacy responsibilities for students and extended the policy's coverage to recent graduates (CPS Board Report 25-1023-PO2).

Responsibility for a secure framework covering classroom AI and machine-learning tools sits jointly with CPS's Department of Information and Technology Services and its Office of Teaching and Learning (CPS Board Report 25-1023-PO2). The same policy assigns the Office of Teaching and Learning responsibility for developing a mandatory, grade-appropriate K-12 AI literacy curriculum, which the policy describes as required under Illinois state law (CPS Board Report 25-1023-PO2). That's the policy's own description of the state requirement, not this article's independent verification of Illinois law, and it's a responsibility assignment, not evidence the curriculum is already running in every classroom.

Before adoption, CPS opened the policy to public comment from Aug. 28 to Oct. 8, 2025 (CPS Board Report 25-1023-PO2). The policy separately guarantees students a learning environment free of discrimination, harassment, abuse, and retaliation (CPS Board Report 25-1023-PO2), a district-wide student-safety principle rather than a spelled-out content rule for any specific AI tool.

Federal guidance allows some AI spending but does not require adoption

CPS's policy sits inside a broader federal signal about AI funding. Last summer, the U.S. Department of Education sent a Dear Colleague Letter to grantees and future grantees on using federal grant funds to support AI in education (U.S. Department of Education).

The letter affirms that AI use is allowable under existing federal education programs, provided it meets applicable statutory and regulatory requirements, and names three examples: AI-based instructional materials, AI-enhanced tutoring, and AI tools for college and career advising (U.S. Department of Education). The letter does not announce a requirement that districts adopt chatbots; it clarifies what existing grant dollars may legally cover.

The same letter names user privacy as a principle for responsible AI adoption, calls for teaching students appropriate AI use alongside existing social media lessons, and stresses engaging parents in decisions about new classroom technology (U.S. Department of Education). It doesn't spell out what vendors must disclose about data retention or model training; that detail depends on the specific contract a district signs.

For vendor evaluation, the National Institute of Standards and Technology published a voluntary Generative AI Profile in mid-2024, updated earlier this year, meant to help organizations build trustworthiness considerations into how they design, develop, use, and evaluate AI systems (NIST, AI RMF Generative AI Profile). It's a general, cross-sector resource, not a K-12-specific rule; applying it to school vendor review is this article's suggestion, not a use NIST specifies.

A practical checklist for districts considering K-12 AI chatbot implementation

None of the sources above lay out one required process for adopting a school chatbot. The five questions below combine CPS's documented framework with federal guidance into an editorial checklist. They reflect this article's recommendations, not CPS policy, federal requirements, or legal advice, and a district's own counsel, IT staff, and school board have final say over what applies locally.

1. Who approved this, and for what purpose?

Before students get access, a school or district can document a written educational purpose and name the office that approved it. If no one in the building can name who signed off, that's a reasonable sign the pilot isn't ready.

2. What do the vendor's data terms say?

District legal or privacy staff can request written answers from the vendor on data retention and model training before signing. Whether FERPA, COPPA, or a state student-privacy law applies depends on the tool's specific use and the student data involved, a determination that belongs to district counsel rather than a general checklist like this one.

3. Who is the human fallback?

Teachers and school leaders can name a real person a student is told to contact if a chatbot's answer seems wrong or unsafe. For a tool that touches a grade, a disciplinary flag, or an advising recommendation, many districts choose to keep a qualified person accountable for the final call rather than letting the tool's output stand on its own.

4. Has staff training happened?

Teachers benefit from a published, assignment-level rule before students use the tool, for example permitting brainstorming while restricting submission of AI-generated text as original work. Accessibility belongs in this step too: screen-reader compatibility, multilingual support, and no assumption that every student has home internet or prior AI experience.

5. What will the pilot measure, and when does it end?

A defined end date, a reporting channel, and tracked outcomes, engagement, error rates, reported incidents, and staff workload, give a pilot a way to be judged rather than just run indefinitely. Some districts choose to pause a pilot after a documented data breach or harmful output, while treating low engagement alone as a reason to adjust the tool rather than shut it down. Both are editorial suggestions here, not a sourced safety standard.

What to check before a chatbot reaches a classroom

Before a district signs a vendor agreement, or before a parent lets a student log into a school-provided chatbot, someone should be able to point to where written answers to those five questions live: a board policy, a vendor contract, or an internal procurement file.

A policy page like CPS's "AI at CPS" site describes district-wide rules and responsibilities. It is not the same as a procurement record, which would show whether a specific product was purchased, or a vendor contract, which would show what that product does with student data. Confirming a chatbot deployment requires the second and third document, not just the first.

Parents, teachers, and school leaders can read CPS's Student Acceptable Use policy directly on the district's board policy site and compare it against whatever AI policy their own school or district has published, if one exists. If a school can't identify where its own answers to those five questions are documented, that's the question to bring to the district's technology, privacy, or instructional office before student access begins, not a reason to assume the tool is unsafe or that a rollout is already underway.