Record revision
Updates to this site's reviewed record, not confirmation that the university changed its policy.
Tracker publication time is not a university policy effective date.
Records involved: 45 (29 corrected / adjusted · 16 added)
Approved tools · Data safety · 8 more topics
Comparable old and new text is unavailable.
Cornell's IT guidelines endorse a flexible framework in which faculty and instructors can choose to prohibit, allow with attribution, or encourage generative AI use in education.
Cornell's IT guidelines endorse a flexible framework in which faculty and instructors can choose to prohibit, allow with attribution, or encourage generative AI use in education.
Under Code I.B.2, using, obtaining or providing unauthorized assistance on examinations, papers or other academic work violates academic integrity. Unauthorized assistance includes AI aid not explicitly permitted by the instructor.
Cornell’s undergraduate guide tells students to ask their professor before using AI when permission is unclear. Permitted AI use must cite the tool, quote verbatim output, acknowledge paraphrases and follow instructor-specific citation expectations; policies differ by course.
The Academic Integrity Code was updated July 27, 2026 and is effective August 24, 2026. Issues related to courses prior to that date use the previous Code. Graduate students are subject to the Code but use graduate-specific procedures; Law and Veterinary honor codes have separate governance.
Code I.C.1 prohibits assistance or information during in-class exams unless given in the exam or by the proctor, and restricts take-home examination consultation with other persons. Students must understand the examination conditions; I.B.2 separately covers unauthorized AI assistance.
Comparable old and new text is unavailable.
When generative AI use is permitted in a course, Cornell advises instructors to clarify expectations for documentation and attribution, including citing the AI tool creator (e.g., OpenAI for ChatGPT) when directly quoting AI-generated text in both in-text citations and reference lists.
When generative AI use is permitted in a course, Cornell advises instructors to clarify expectations for documentation and attribution, including citing the AI tool creator (e.g., OpenAI for ChatGPT) when directly quoting AI-generated text in both in-text citations and reference lists.
In Cornell Communication PhD A/B exams, all work must be the student’s own: students may not use GenAI to draft or write A-exam papers or the dissertation. AI-assisted language/grammar improvements must be disclosed, including prompts, suggestions and revisions. Students must discuss AI use in formative, reporting and writing phases with their committee; the policy recommends a reflexive appendix.
Cornell AI Platform is listed for Cornell University in an official university AI tools source. Derived availability: conditionally allowed. Derived endorsement type: officially endorsed.
Cornell AI Platform has the stated Cornell access and data conditions; its listing does not authorize coursework or examinations. Limited pilot collaborators; contact Innovation Hub. Moderate Risk and below per tool entry.
Microsoft Copilot Basic is listed for Cornell University in an official university AI tools source. Derived availability: allowed. Derived endorsement type: officially endorsed.
Microsoft Copilot Basic has the stated Cornell access and data conditions; its listing does not authorize coursework or examinations. Eligible faculty/staff/students aged 18+; use Cornell authentication. Only Low Risk data; public search has limited privacy.
Cornell's committee report states that any information educators are obligated to keep private under FERPA or HIPAA should not be shared with generative AI tools or uploaded to third-party AI vendors.
Cornell's 2023 education committee report warns against disclosing FERPA/HIPAA-protected information to third-party generative AI services where submitted data may be used for training. This is the report's third-party data-protection context, not a blanket prohibition covering every institutionally approved deployment.
Cornell's committee report states that original research or content owned by Cornell University, students, or employees should not be uploaded to AI tools, as it can become part of the AI tool's training data.
Cornell's 2023 education committee report cautions against uploading original research or Cornell-, student- or employee-owned content to the third-party generative AI tools it discusses, because submitted material may become training data. Institutionally controlled tools require their own data safeguards and authorizations.
Comparable old and new text is unavailable.
Cornell's IT guidelines prohibit entering any confidential, proprietary, federally or state regulated, or otherwise sensitive or restricted Cornell information into public generative AI tools, as such information becomes public and may be stored and used by anyone.
Cornell's IT guidelines prohibit entering any confidential, proprietary, federally or state regulated, or otherwise sensitive or restricted Cornell information into public generative AI tools, as such information becomes public and may be stored and used by anyone.
Adobe Firefly is listed for Cornell University in an official university AI tools source. Derived availability: conditionally allowed. Derived endorsement type: institutionally licensed or procured.
Adobe Firefly has the stated Cornell access and data conditions; its listing does not authorize coursework or examinations. Adobe licence required; contact Hub/local IT to confirm entitlement.
Zoom AI Companion has the stated Cornell access and data conditions; its listing does not authorize coursework or examinations. Security/privacy evaluation is ongoing; access remains unconfirmed.
Cornell's IT guidelines state that use of generative AI in academic research is governed by the Cornell University Task Force Report 'Generative AI in Academic Research: Perspectives and Cultural Norms' (December 2023).
Cornell IT refers researchers to the December 2023 task-force report for perspectives and practical guidance on generative AI in academic research. The report explicitly describes itself as internal advice rather than binding rules; applicable research policies and unit-specific requirements must be assessed separately.
Accepting Responsibility requires the instructor to believe evidence meets the clear-and-convincing threshold and limits the penalty to the coursework charged. A prior responsibility agreement or guilt finding makes a student ineligible. Eligible students have three business days to choose and, after choosing responsibility, three business days to reconsider; silence selects the ordinary hearing route.
Accepting Responsibility is voluntary for both instructor and student; either may choose Primary Hearing and Appeals instead. It is not a guilt finding, but requires a limited coursework penalty and educational workshop. Choosing it waives appeals to the AIHB and dean; procedural questions may go to the Dean of Faculty.
A student accepting responsibility must complete the in-person educational workshop and any outside work within two weeks. Default allows the faculty member to begin a Primary Hearing and Appeals process without being bound by the earlier penalty and makes the student ineligible for future Accepting Responsibility.
At Board hearings students may have an advisor and relevant witnesses, present their case and challenge charges or evidence; an advisor may assist presentation and questioning. The Board finds guilt only on clear and convincing evidence. Law/Veterinary honor-code and graduate-specific procedures must be read separately.
A student may seek Board review for unfair/improper procedure, a contested finding or excessive penalty. The individual seeking review notifies the chair of the faculty member’s college Hearing Board within ten working days of the Primary Hearing; the chair may allow a good-cause exception. Board review is de novo; an increased penalty on student-initiated review should be exceptional.
A student may appeal the Board decision in writing to their college dean, normally within four weeks of notification; the dean may grant a good-cause exception. Guilt appeals concern new outcome-relevant evidence or prejudicial Board procedure; penalty appeals explain inappropriateness. The dean’s decision cannot be appealed under this Code route.
Academic integrity findings use a clear-and-convincing standard. Faculty investigate apparent violations and may discuss a student’s process, information sources and techniques before declining a case, notifying a Primary Hearing, or offering Accepting Responsibility.
Code I.B.3 identifies fabricated or non-existent citations, including AI-generated false citations, as violations. Students are responsible for the existence and accuracy of every source they cite.
Under II.D, the charge includes notice of a Primary Hearing with at least one week’s notice, which both parties may agree to shorten, and notice of Respondents’ Codes Counselor availability. The student, faculty member and independent witness attend; the student may bring an advisor and witnesses and respond with evidence.
After hearing the student, the instructor may dismiss the charge or find guilt only on clear and convincing evidence, and may impose suitable grade punishment including course failure. Primary Hearing findings are reported to the registry and shared with the student. Penalties beyond grades involve Board/dean processes; no automatic penalty follows AI use or detector output.
Comparable old and new text is unavailable.
Cornell's Center for Teaching Innovation recommends that faculty discuss course policies and expectations around the use of generative AI tools with their students and clearly communicate when and in what ways use of such tools is permitted or not.
Cornell's Center for Teaching Innovation recommends that faculty discuss course policies and expectations around the use of generative AI tools with their students and clearly communicate when and in what ways use of such tools is permitted or not.
Comparable old and new text is unavailable.
Cornell's committee report does not recommend the use of generative AI for summative evaluation or grading of student work, stating that evaluation and grading is among the most important tasks entrusted to faculty.
Cornell's committee report does not recommend the use of generative AI for summative evaluation or grading of student work, stating that evaluation and grading is among the most important tasks entrusted to faculty.
Comparable old and new text is unavailable.
Cornell recommends that faculty clearly communicate their generative AI policies in their syllabus, in assignment instructions, and verbally in class to support student learning and reduce academic integrity violations.
Cornell recommends that faculty clearly communicate their generative AI policies in their syllabus, in assignment instructions, and verbally in class to support student learning and reduce academic integrity violations.
Comparable old and new text is unavailable.
Cornell's IT guidelines state that use of generative AI for administrative purposes must comply with the Cornell Generative AI in Administration Task Force Report (January 2024).
Cornell's IT guidelines state that use of generative AI for administrative purposes must comply with the Cornell Generative AI in Administration Task Force Report (January 2024).
Comparable old and new text is unavailable.
Cornell developed a set of course policy icons through the GenAI Advisory Council to help instructors clearly and consistently communicate AI use expectations to students in syllabi and assignment instructions.
Cornell developed a set of course policy icons through the GenAI Advisory Council to help instructors clearly and consistently communicate AI use expectations to students in syllabi and assignment instructions.
Comparable old and new text is unavailable.
Cornell describes measures faculty may use to evaluate potential AI-related academic integrity concerns, including requiring students to verify citations and references, requesting verification of references or methods, and informing students that they should expect to verbally explain submitted work.
Cornell describes measures faculty may use to evaluate potential AI-related academic integrity concerns, including requiring students to verify citations and references, requesting verification of references or methods, and informing students that they should expect to verbally explain submitted work.
Cornell's committee report recommends that the Code of Academic Integrity be updated with clear and explicit language on the use of generative AI, indicating that individual faculty have authority to determine when AI use is prohibited, attributed, or encouraged.
Cornell's 2023 education committee recommended explicit AI language in the Academic Integrity Code and faculty authority over course use. This is a historical recommendation; the Code effective August 24, 2026 now expressly addresses unauthorized AI assistance and fabricated citations.
Comparable old and new text is unavailable.
Cornell's committee report recommends three policy approaches for generative AI use: prohibit GAI where it interferes with foundational learning, allow with attribution where it supports higher-level thinking, and encourage use where it enables exploration and creative thinking.
Cornell's committee report recommends three policy approaches for generative AI use: prohibit GAI where it interferes with foundational learning, allow with attribution where it supports higher-level thinking, and encourage use where it enables exploration and creative thinking.
Comparable old and new text is unavailable.
Cornell's committee report discourages the use of automatic detection algorithms for academic integrity violations using generative AI, stating they cannot decisively provide evidence and could lead to unfairly identifying violations, including bias against non-native speakers.
Cornell's committee report discourages the use of automatic detection algorithms for academic integrity violations using generative AI, stating they cannot decisively provide evidence and could lead to unfairly identifying violations, including bias against non-native speakers.
Comparable old and new text is unavailable.
Cornell's IT guidelines state that users are accountable for their work regardless of the tools used to produce it, and when using generative AI tools must always verify information for errors and biases and exercise caution to avoid copyright infringement.
Cornell's IT guidelines state that users are accountable for their work regardless of the tools used to produce it, and when using generative AI tools must always verify information for errors and biases and exercise caution to avoid copyright infringement.
Comparable old and new text is unavailable.
Cornell has established seven core principles for generative AI in education: integrity of the faculty-student relation, commitment to experimentation and evidence, centrality of faculty judgment, responsiveness to student needs, recognition of both AI goods and harms, respect for institutional and disciplinary heterogeneity, and renewal of Cornell's core mission and values.
Cornell has established seven core principles for generative AI in education: integrity of the faculty-student relation, commitment to experimentation and evidence, centrality of faculty judgment, responsiveness to student needs, recognition of both AI goods and harms, respect for institutional and disciplinary heterogeneity, and renewal of Cornell's core mission and values.
Comparable old and new text is unavailable.
Cornell established the GenAI Education Working Group in spring 2024 with all-college representation including faculty, staff, and students, as the central body for developing new ideas, policies, and practices around generative AI in the classroom.
Cornell established the GenAI Education Working Group in spring 2024 with all-college representation including faculty, staff, and students, as the central body for developing new ideas, policies, and practices around generative AI in the classroom.
Comparable old and new text is unavailable.
Cornell's 'AI+AI' initiative aims to strengthen and update academic integrity procedures to better reflect the presence of generative AI, including better models for attributing student use of GenAI and development of evidentiary standards for adjudicating GenAI-related violations.
Cornell's 'AI+AI' initiative aims to strengthen and update academic integrity procedures to better reflect the presence of generative AI, including better models for attributing student use of GenAI and development of evidentiary standards for adjudicating GenAI-related violations.
Comparable old and new text is unavailable.
Cornell developed seven standardized AI course policy icons (ANY-AI, AT, UA, PP, AS, ER, AI-FREE) to help instructors clearly communicate AI use expectations in syllabi and assignments, which can be combined for nuanced policies.
Cornell developed seven standardized AI course policy icons (ANY-AI, AT, UA, PP, AS, ER, AI-FREE) to help instructors clearly communicate AI use expectations in syllabi and assignments, which can be combined for nuanced policies.
Comparable old and new text is unavailable.
Cornell's AI course policy icons include a 'PP' (Privacy Protecting) icon indicating that generative AI use is permitted but no copyrighted or proprietary class materials should be uploaded unless otherwise specified.
Cornell's AI course policy icons include a 'PP' (Privacy Protecting) icon indicating that generative AI use is permitted but no copyrighted or proprietary class materials should be uploaded unless otherwise specified.
Comparable old and new text is unavailable.
Cornell considers generative AI literacy essential for both students and faculty, defining it as the ability to understand, evaluate, and critically engage with generative AI technologies.
Cornell considers generative AI literacy essential for both students and faculty, defining it as the ability to understand, evaluate, and critically engage with generative AI technologies.
Comparable old and new text is unavailable.
Cornell holds students responsible for verifying the accuracy of AI-generated output and references when AI use is allowed for an assignment.
Cornell holds students responsible for verifying the accuracy of AI-generated output and references when AI use is allowed for an assignment.
Comparable old and new text is unavailable.
Cornell does not recommend using automatic AI detection algorithms for academic integrity violations, citing their unreliability and inability to provide definitive evidence, and the risk of wrongly accusing students.
Cornell does not recommend using automatic AI detection algorithms for academic integrity violations, citing their unreliability and inability to provide definitive evidence, and the risk of wrongly accusing students.
Comparable old and new text is unavailable.
Cornell provides sample syllabus language for an AI-FREE policy that prohibits all generative AI tools, including tools that help reorganize and edit written work, to ensure development of foundational concepts and skills.
Cornell provides sample syllabus language for an AI-FREE policy that prohibits all generative AI tools, including tools that help reorganize and edit written work, to ensure development of foundational concepts and skills.
Comparable old and new text is unavailable.
Cornell provides sample syllabus language for an AS-UA policy where AI use is generally discouraged but permitted for select assignments with proper attribution, requiring students to cite the AI tool creator.
Cornell provides sample syllabus language for an AS-UA policy where AI use is generally discouraged but permitted for select assignments with proper attribution, requiring students to cite the AI tool creator.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
Comparable old and new text is unavailable.
