Révision du registre
Mises à jour de nos données vérifiées ; elles ne confirment pas un changement de politique de l’université.
La date de publication du registre n’est pas la date d’entrée en vigueur de la politique.
Éléments concernés : 39 (30 corrigés / ajustés · 9 ajoutés)
Règles d’utilisation des outils d’IA · Enseignement · 7 autres thèmes
Textes ancien et nouveau non comparables.
UTokyo states that submitting AI-generated answers by copy-pasting them entirely provides no learning effect and should basically not be permitted.
UTokyo states that submitting AI-generated answers by copy-pasting them entirely provides no learning effect and should basically not be permitted.
Textes ancien et nouveau non comparables.
UTokyo states it is unacceptable to present AI-generated text as one's own when submitting class assignments.
UTokyo states it is unacceptable to present AI-generated text as one's own when submitting class assignments.
Textes ancien et nouveau non comparables.
UTokyo policy states that even when AI use is allowed, simply copying and pasting AI-generated answers should not be accepted as it provides no learning effect.
UTokyo policy states that even when AI use is allowed, simply copying and pasting AI-generated answers should not be accepted as it provides no learning effect.
When students submit reports that utilized generative AI tools, course instructors are officially recommended to have students clearly specify which generative AI tools were used; specific attribution format and disclosure requirements remain at faculty discretion.
Textes ancien et nouveau non comparables.
UTokyo leaves decisions on AI tool use in individual classes to faculty, who should decide based on maximizing educational effectiveness rather than following a blanket university rule.
UTokyo leaves decisions on AI tool use in individual classes to faculty, who should decide based on maximizing educational effectiveness rather than following a blanket university rule.
Textes ancien et nouveau non comparables.
UTokyo warns faculty not to over-rely on AI detection tools, as they are insufficient evidence of inappropriate student AI use given the rapid evolution of generative tools.
UTokyo warns faculty not to over-rely on AI detection tools, as they are insufficient evidence of inappropriate student AI use given the rapid evolution of generative tools.
Textes ancien et nouveau non comparables.
UTokyo leaves the decision on whether to allow AI tools and the conditions for their use to the discretion of individual instructors, varying by class.
UTokyo leaves the decision on whether to allow AI tools and the conditions for their use to the discretion of individual instructors, varying by class.
Textes ancien et nouveau non comparables.
UTokyo advises faculty to test their own assignments with generative AI tools to understand how well AI can complete them, and use this understanding to inform assessment design.
UTokyo advises faculty to test their own assignments with generative AI tools to understand how well AI can complete them, and use this understanding to inform assessment design.
UTokyo requires instructors to clearly state their AI stance per class/assignment, and when allowing AI use, to explain risks: information leakage, data concentration, copyright concerns, and potential bias.
UTokyo requires course instructors to clearly designate permitted AI use, which effective September 29, 2026 must be configured in UTOL using standardized labels (T0–T5 or Other).
Textes ancien et nouveau non comparables.
UTokyo warns faculty that AI detection tools should not be over-relied upon, as their judgments are insufficient evidence of inappropriate student AI use.
UTokyo warns faculty that AI detection tools should not be over-relied upon, as their judgments are insufficient evidence of inappropriate student AI use.
Effective September 29, 2026, setting 'Permitted AI Use' is mandatory for all assignment creation in UTOL (UTokyo Online Learning), requiring course instructors to designate a standardized permission level or provide explicit custom instructions.
UTOL assignments classify permitted generative AI use into six standardized ranges (T0 No AI Use; T1 questions/research only; T2 discussion/idea-bouncing; T3 feedback; T4 drafting; T5 creative use) or 'Other' with required written explanation.
The University of Tokyo educational AI framework classifies typical AI use scenarios into five sequential categories (S1 questions/research; S2 discussion/idea-bouncing; S3 feedback; S4 drafting; S5 creative use) mapping directly to UTOL assignment permission labels T0–T5.
Generative AI tool licensing and technical availability never grant automatic coursework assessment permission; use in assignments or exams is strictly governed by the course instructor's syllabus and UTOL assignment settings.
Textes ancien et nouveau non comparables.
UTokyo requires instructors who allow AI use to explain associated risks to students: information leakage, data concentration in few companies, copyright concerns, and potential bias.
UTokyo requires instructors who allow AI use to explain associated risks to students: information leakage, data concentration in few companies, copyright concerns, and potential bias.
Textes ancien et nouveau non comparables.
UTokyo instructs faculty not to input exam questions directly into generative AI tools, as exams are highly confidential documents.
UTokyo instructs faculty not to input exam questions directly into generative AI tools, as exams are highly confidential documents.
Textes ancien et nouveau non comparables.
UTokyo warns students to never input confidential information, personal information, or unpublished research results into AI tools, as the information might be leaked or used for AI training.
UTokyo warns students to never input confidential information, personal information, or unpublished research results into AI tools, as the information might be leaked or used for AI training.
Under official faculty guidance and student notices, students and faculty are cautioned against information leakage: students must never input confidential information, personal information, or unpublished research results into AI tools due to potential leakage or model training use, and faculty are instructed that examination questions, being highly confidential documents, must basically (基本的に) not be input into language-generating AI in unchanged form. The separate 'Security Guidelines for Generative AI Use' (March 12, 2024) require internal UTokyo Account authentication and are held as an unverified private source.
UTokyo has established security guidelines for generative AI use (March 2024), referenced in the AI service policy.
Historical March2024 interim service notice (not a current comprehensive service inventory): The public March27,2024 notice links to Security Guidelines for Generative AI Use dated March12,2024. Only the reference/title was captured; its authenticated full text and detailed requirements remain unverified.
Under Rule No. 79 (東京大学科学研究行動規範委員会規則), formal scientific research misconduct investigations establish procedural timelines qualified in principle (原則として): the preliminary inquiry results must be reported to the Committee Chair in principle within 30 days of receiving an allegation (Art. 8(5)); the Investigation Committee commences formal investigation in principle within 30 days of the decision to investigate (Art. 9(2)); the Committee makes its adjudication determination in principle within 150 days of starting formal investigation (Art. 11(1)); respondent defense opportunity is granted in principle within 14 days of receiving draft determination notification (Art. 11(7)); and a respondent against whom misconduct was determined may file an objection/appeal to the Committee Chair in principle within 30 days of receiving the determination notice under Art. 11(11) (Art. 12(1)). These clocks apply exclusively to scientific research misconduct proceedings and do not govern undergraduate or generic coursework discipline.
Textes ancien et nouveau non comparables.
Cisco AI アシスタント is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Cisco AI アシスタント is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Gemini is listed for The University of Tokyo in an official university AI tools source. Derived availability: allowed. Derived endorsement type: institutionally licensed or procured.
Google Gemini is provided to all members of The University of Tokyo with ECCS Cloud Email accounts under institutional data protection; coursework use is governed by instructor assignment settings on UTOL.
Microsoft 365 Copilot Chat is listed for The University of Tokyo in an official university AI tools source. Derived availability: allowed. Derived endorsement type: institutionally licensed or procured.
Historical March2024 interim service notice (not a current comprehensive service inventory): Microsoft 365 Copilot Chat is provided to all members of The University of Tokyo with UTokyo Accounts with commercial data protection; coursework use is governed by instructor assignment settings on UTOL.
Textes ancien et nouveau non comparables.
Copilot in Teams is listed for The University of Tokyo in an official university AI tools source. Derived availability: conditionally allowed. Derived endorsement type: institutionally licensed or procured.
Copilot in Teams is listed for The University of Tokyo in an official university AI tools source. Derived availability: conditionally allowed. Derived endorsement type: institutionally licensed or procured.
Textes ancien et nouveau non comparables.
Slack AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Slack AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Textes ancien et nouveau non comparables.
Slido AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Slido AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Textes ancien et nouveau non comparables.
Webex AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Webex AI is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Zoom AI Companion is listed for The University of Tokyo in an official university AI tools source. Derived availability: not mentioned. Derived endorsement type: institutionally licensed or procured.
Zoom AI Companion is provided to eligible members of The University of Tokyo under institutional Zoom licenses; institutional access does not grant assessment permission.
Textes ancien et nouveau non comparables.
University of Tokyo will not uniformly prohibit generative AI tools like ChatGPT in education; instead, it actively explores their potential while continuing dialogue on practical knowledge and long-term impact.
University of Tokyo will not uniformly prohibit generative AI tools like ChatGPT in education; instead, it actively explores their potential while continuing dialogue on practical knowledge and long-term impact.
UTokyo does not uniformly prohibit generative AI tools; it actively explores their potential in education and research while providing practical usage guidance.
Historical March2024 interim service notice (not a current comprehensive service inventory): UTokyo does not uniformly prohibit generative AI tools; it actively explores their potential in education and research while providing practical usage guidance.
Textes ancien et nouveau non comparables.
UTokyo does not enforce a blanket prohibition on generative AI tools; it actively explores their potential and provides practical guidance.
UTokyo does not enforce a blanket prohibition on generative AI tools; it actively explores their potential and provides practical guidance.
Textes ancien et nouveau non comparables.
University of Tokyo will not uniformly prohibit generative AI tools like ChatGPT in educational settings, per official policy signed by the Executive Vice President.
University of Tokyo will not uniformly prohibit generative AI tools like ChatGPT in educational settings, per official policy signed by the Executive Vice President.
UTokyo provides Gemini and Microsoft 365 Copilot Chat as university-wide contracted AI chat services; there is no university-wide ChatGPT contract.
The captured utelecon catalogue names Gemini and Microsoft 365 Copilot Chat as the two main university-contracted chat AI services. This listing does not establish that no other institutional contract exists.
Effective February 24, 2026, Gemini generative AI features are integrated into Google Docs, Google Slides, and Google Forms for U-Tokyo users assigned Google Workspace for Education Plus licenses; tool presence does not authorize coursework drafting unless permitted by the instructor.
Zoom AI Companion features (meeting summaries, in-meeting queries, smart recording highlights) are provided under U-Tokyo's campus-wide Zoom license, with meeting hosts controlling activation; institutional access does not grant coursework or exam permission.
UTokyo centers its university-wide generative AI service on Microsoft Copilot, leveraging existing Microsoft contracts for low operational and cost burden.
Historical March2024 interim service notice (not a current comprehensive service inventory): UTokyo centers its university-wide generative AI service on Microsoft Copilot, leveraging existing Microsoft contracts for low operational and cost burden.
Textes ancien et nouveau non comparables.
UTokyo provides AI features embedded in multiple platforms (Zoom AI Companion, Slack AI, Cisco AI Assistant, Slido AI, Copilot in Teams) through university-wide contracts.
UTokyo provides AI features embedded in multiple platforms (Zoom AI Companion, Slack AI, Cisco AI Assistant, Slido AI, Copilot in Teams) through university-wide contracts.
Textes ancien et nouveau non comparables.
UTokyo's utelecon portal maintains a Policies and Guidelines index page linking to BYOD policy, generative AI information, and AI tools notifications.
UTokyo's utelecon portal maintains a Policies and Guidelines index page linking to BYOD policy, generative AI information, and AI tools notifications.
Textes ancien et nouveau non comparables.
UTokyo warns that using AI-generated results as-is may entail future legal risks related to copyright and design rights.
UTokyo warns that using AI-generated results as-is may entail future legal risks related to copyright and design rights.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
Textes ancien et nouveau non comparables.
