Open, evidence-backed AI policy records for public reuse.
New York City, United States
Columbia University is listed as QS 2026 rank =38. Columbia University has 16 source-backed AI policy claim records from 8 official source attributions. The public record preserves original-language evidence snippets, source URLs, snapshot hashes, confidence, and review state.
v1 public contract
Columbia University is listed as QS 2026 rank =38. Columbia University has 16 source-backed AI policy claim records from 8 official source attributions. The public record preserves original-language evidence snippets, source URLs, snapshot hashes, confidence, and review state.
Candidate claims are source-backed records pending review. They are not final policy conclusions and are not legal or academic integrity advice.
11 reviewed public claim
Academic Integrity
Original evidence
Evidence 1Use of Generative AI is prohibited in (a) any exam or final paper or (b) for aid in drafting any part of work submitted for credit, even if the use is fully documented.
Ai Tool Treatment
Original evidence
Evidence 1Columbia University provides access to HIPAA-compliant versions of OpenAI's ChatGPT and Microsoft Copilot, enabling our workforce to leverage these AI tools responsibly and compliantly.
Privacy
Original evidence
Evidence 1Sensitive Data: Permitted only on the ChatGPT Education, approved Microsoft CoPilot platforms, and CU CHAT when used with compliant models. Research protocol use requires IRB, and TRAC/ACORD approval.
Teaching
Original evidence
Evidence 1Individual instructors can, and indeed are encouraged to, tailor their own more permissive policies, so long as their policies are stated in writing in the syllabus.
Ai Tool Treatment
Original evidence
Evidence 1Students may use Generative AI to aid in studying, brainstorming, or to identify typographical errors.
Privacy
Original evidence
Evidence 1All uses of Generative AI must comply with University policy protecting confidential and personal information. By default, all text you enter into Generative AI tools is retained, used for training, and potentially outputted to other users.
Ai Tool Treatment
Original evidence
Evidence 1At this time (March 2026), the following AI Chat services offered through CUIT are not approved for use with Sensitive data: Google Gemini, NotebookLM, Anthropic Claude.
Security Review
Original evidence
Evidence 1For locally installed AI models (i.e. LLM, NLP, ML), a formal IT Risk Assessment review is required before deployment to evaluate potential security, privacy, and compliance risks.
Teaching
Original evidence
Evidence 1Example 1: No Generative AI Use Permitted. Students are prohibited from using generative Artificial Intelligence (AI) tools to complete coursework or assignments for this class.
Teaching
Original evidence
Evidence 1Example 5: Generative AI tools are generally permitted in this course for research and completion of assignments, provided that all AI-generated content is clearly attributed as such in the student's work.
Teaching
Original evidence
Evidence 1Students with disabilities are eligible for reasonable accommodations to permit them equal access to Teachers College programs and services (which include classes and coursework). If you are a student registered with OASID with a generative AI accommodation, please speak with me directly about your needs.
5 machine or needs-review claim
Candidate claims are not final policy conclusions. They preserve source URL, source snapshot hash, evidence, confidence, and review state so the record can be audited before review.
Source Status
This claim is held for review because the evidence or classification needs another pass.
Original evidence
Evidence 1Columbia University AI Guidelines hub from CTL. Links to Provost policy, academic integrity resources, CUIT resources, best practices for responsible AI use. Central hub for Columbia AI resources.
Source Status
This claim is held for review because the evidence or classification needs another pass.
Original evidence
Evidence 1This Generative AI policy governs the use of Generative AI tools by staff, faculty, students, and researchers. Covers information security, data privacy, copyright, academic integrity, bias.
Research
This claim is held for review because the evidence or classification needs another pass.
Original evidence
Evidence 1Researchers must avoid uploading, or using as input, any unpublished research data or other Confidential Information into a Generative AI tool.
Academic Integrity
This claim is held for review because the evidence or classification needs another pass.
Original evidence
Evidence 1As a general rule, students should disclose to faculty if they are using generative AI platforms and in what manner they are using them in coursework.
Source Status
This claim is held for review because the evidence or classification needs another pass.
Original evidence
Evidence 1CUIT university-wide best practices for responsible AI use. Referenced from ai.ctl.columbia.edu hub. Applies to faculty, students, researchers, and staff.
8 source attribution
it.cuimc.columbia.edu
etc.cuit.columbia.edu
law.columbia.edu
ai.ctl.columbia.edu
provost.columbia.edu
provost.columbia.edu
students.business.columbia.edu
tc.columbia.edu