Policy record and evidence correction

Add ANU, Edinburgh, and CUHK policy records with verified claims, exact presentation patches, and restored CUHK analysis dimensions.

Historical correction: counts and claims reflect this version, not the latest selected record.

2026-10-21T08:35:00.000Z · v6-target-r7-candidate-20261021

• The University of New South Wales (UNSW Sydney) · 40 reviewed records in this revision · View policy page
  • Added records: 0; Corrected records: 0; Held / excluded from current conclusions: 0

Previous V6 release history

Record revision, scope and original evidence

Original evidence and audience, course and unit scopes are preserved. Tracker correction time is not a policy effective date.

The records below reflect this historical version, not the current policy record or current review conclusions.

162f56fa911147609cee1e855195b0f2db3311a172b36b4d849543ae2ebe422d

{
  "addedClaimIds": [],
  "updatedClaimIds": [],
  "heldClaimIds": []
}

• claim-unsw-sydney-2 — UNSW's student guidance lists six Levels of AI Assistance for Term 3 2026, Semester 2 2026 and Hexamester 5 2026: No Assistance, Simple Editing Assistance, Planning or Design Assistance, Assistance with Attribution, Generative AI Software-based Assessments, and Not Applicable. For Hexamester 6 2026 and Summer Term 2027, the page specifies three AI Assessment Modes: Can't Use AI, Can Use AI and Must Use AI. Students should use the framework for their teaching period and check each assessment's instructions.
• claim-unsw-sydney-13 — The retained Staff Teaching Gateway evidence describes categories of permitted AI use to support assessment design. Current student guidance distinguishes the six Levels of AI Assistance for Term 3, Semester 2 and Hexamester 5 2026 from three AI Assessment Modes for Hexamester 6 2026 and Summer Term 2027.
• claim-unsw-sydney-3 — Under UNSW's 'No Assistance' level, students are not permitted to use any generative AI tools, software, or service to search for or generate information or answers.
• claim-unsw-sydney-4 — At UNSW, the unauthorised or unacknowledged use of AI in assessments is classified as cheating and considered student misconduct under the Code of Conduct and Values.
• cl-the-university-of-new-south-wales-unsw-sydney-microsoft-copilot-150 — Microsoft Copilot Chat is listed for The University of New South Wales (UNSW Sydney) in an official university AI tools source. Derived availability: allowed. Derived endorsement type: institutionally licensed or procured.
• claim-unsw-sydney-10 — UNSW has published six AI principles: AI use must benefit UNSW and society; be equitable and respectful of human rights; be trustworthy, safe, and reliable; be transparent; be identifiable, explainable, and contestable; and be secure and resilient.
• claim-unsw-sydney-11 — UNSW's first key principle for AI in assessment requires staff to be honest and transparent about the use of any AI tool where it would reasonably be expected that use of the tool would be disclosed.
• claim-unsw-sydney-12 — UNSW's second key principle for AI in assessment requires that any AI-based output must be reviewed with all due diligence before being released or relied upon, particularly to avoid bias and factual errors.
• claim-unsw-sydney-15 — UNSW states that as a rule, markers must not use AI platforms for marking or grading student work.
• claim-unsw-sydney-16 — UNSW only authorises the use of Turnitin's AI Writing Detection Tool for detecting improper AI use in student work; UNSW IT has not approved other detection tools due to privacy and accuracy concerns.
• claim-unsw-sydney-18 — UNSW's formal Student Plagiarism Management Procedure uses the levels Poor scholarship, Minor, Moderate, Significant and Serious, with generative AI misuse assessed according to the applicable rubric. Poor scholarship and Minor/Moderate/Significant cases are managed by the School; Serious cases must be referred to the Conduct & Integrity Office (CIO) as potential student misconduct. For Minor plagiarism, the rubric says a mark reduction should not exceed 10% of the maximum grade available. Listed outcomes for Serious cases include failing the work or course, suspension and permanent exclusion; these are possible outcomes, not automatic consequences of every unauthorised AI use. Actions outside the outcomes available for a level require CIO approval, and repeated instances may also be handled as student misconduct.
• claim-unsw-sydney-22 — UNSW requires that students must follow assessment instructions regarding AI use, that what students present must be their own work, and that students must acknowledge all sources including AI as an external source using proper referencing.
• claim-unsw-sydney-24 — UNSW College students may only use AI for an assessment if the instructions in the Course Outline permit its use.
• claim-unsw-sydney-27 — UNSW College requires that if AI is permitted for an assessment, students must acknowledge and reference the AI tool using the referencing style for their course and program (e.g., APA Style Referencing 7th edition).
• claim-unsw-sydney-5 — UNSW penalties for unauthorised AI use in assessments may include a fail grade, a mark of zero for the course, suspension, or permanent exclusion.
• cl-unsw-sydney-mentor-ai-20260924 — Mentor AI is approved by UNSW College for students to practice writing, listening, and speaking skills within English Language Courses (ELICOS).
• claim-unsw-sydney-1 — UNSW advises students given permission to use AI in assessments to always clearly acknowledge that use when planning, designing, or writing their assessment.
• claim-unsw-sydney-14 — UNSW requires that when ChatGPT or other forms of GenAI are accepted as part of an assessment, academics must ensure the tools are easily accessible for all students with no physical, geographical, or financial restrictions.
• claim-unsw-sydney-17 — UNSW's teacher guidance describes preliminary checks and consultation for suspected AI writing. Under the formal Student Plagiarism Management Procedure, the Course Authority or SSIA should notify the student in writing of detected Poor scholarship/Minor/Moderate/Significant plagiarism and proposed academic action; the student has 5 business days to respond. For these School-level outcomes, an internal review must be emailed to the Head of School or delegate within 20 business days of the Notification of Outcome under clause 4.3(a). Review is limited to procedural fairness, not the merits. The reviewer must not have participated in the initial decision; the student is to be notified of the review outcome within 10 business days of receipt of the review request. Serious cases follow the separate CIO process; these School-level deadlines must not be presented as deadlines for every serious-misconduct case.
• claim-unsw-sydney-23 — UNSW states that students are entitled to use generative AI to the extent specified by course instructions or university rules, and that assessment instructions should set out acceptable use of AI tools.
• claim-unsw-sydney-25 — UNSW College recommends Microsoft Copilot as the only generative AI platform for staff and students, as it provides commercial data protection and is the only platform recommended where generative AI use is allowed in coursework and assessment.
• claim-unsw-sydney-6 — UNSW provides academics with access to Turnitin's AI detection tool for assessments submitted through Moodle Turnitin Assignment or Inspera, but notes this is not always conclusive evidence of improper AI use.
• claim-unsw-sydney-7 — UNSW classifies Grammarly, Quillbot, and translation tools such as Google Translate, DeepL, and Baidu Translate as forms of generative AI.
• claim-unsw-sydney-8 — UNSW advises students not to include any personal or sensitive information in AI prompts, including addresses, names, emails, zID, or intellectual property, and recommends using Microsoft Copilot with a UNSW account for data privacy.
• claim-unsw-sydney-20 — UNSW has developed an AI capability framework for teaching staff covering common knowledge (what is AI, prompts, evaluating outputs), AI governance, ethical and responsible use, assessment and learning, and technology (Microsoft Copilot, Adobe Firefly, Scite.ai).
• claim-unsw-sydney-26 — UNSW activated Microsoft Copilot with Commercial Data Protection for all staff and students with a zID in May 2024, providing a secure platform where sensitive information is stored and accessed only by authorised staff.
• claim-unsw-sydney-28 — UNSW College permits AI tools for refining writing only when an AI assessment icon indicates it is allowed, for purposes including refining spelling and grammar, revising style and tone, improving clarity, improving vocabulary, checking similarity, assisting proofreading, and self-marking against a rubric.
• claim-unsw-sydney-9 — UNSW warns students that generative AI tools may produce invalid or unreliable information ('fabrications' or 'hallucinations') and advises them to verify the accuracy of AI outputs.
• claim-unsw-sydney-19 — UNSW has an AI leadership group that serves as a clearinghouse for critical issues and strategic initiatives and plays a pivotal role in guiding enterprise decisions on AI.
• claim-unsw-sydney-21 — Under the AI capability framework for teaching staff, UNSW lists Microsoft Copilot, Adobe Firefly, and Scite.ai as adopted AI tools for identification and practical application. The framework provides capability outcomes for staff to explore features and professional applications of generative AI tools including Copilot, ChatGPT, and Firefly to enhance productivity and innovation.
• claim-unsw-sydney-29 — UNSW's AI Leadership Group endorsed and published principles on the Ethical and Responsible Use of Artificial Intelligence at UNSW in 2024, designed to regulate AI use, support a positive approach, and promote world-leading AI research.
• claim-unsw-sydney-maintenance-20260604-1 — UNSW College says students may use AI for an assessment only when the instructions in the Course Outline permit its use.
• claim-unsw-sydney-maintenance-20260604-3 — UNSW College says students permitted to use AI for an assessment must acknowledge and reference the AI tool using the referencing style for their course and program.
• claim-unsw-sydney-maintenance-20260604-4 — UNSW College says students cannot use AI for an assessment marked with the Category 1 AI icon, and using AI when prohibited may be penalised as academic misconduct.
• claim-unsw-sydney-maintenance-20260604-2 — UNSW College recommends Microsoft Copilot with Commercial Data Protection as the generative AI platform for coursework or assessment when generative AI use is allowed.
• claim-unsw-sydney-maintenance-20260604-5 — UNSW College says ELICOS students should speak to their teacher about using AI for language-learning support such as writing prompts, chatbot conversation practice, pronunciation feedback, listening practice, and writing feedback.
• draft-claim-unsw-assessment-procedure-clause-6-1 — Under Section 6.1 of UNSW's Assessment and Feedback Procedure, the course outline will be published at least one week before the start of the teaching period and will specify for each assessment the allowed role of artificial intelligence and, where relevant, how it can be used. The procedure specifies phased applicability across teaching periods: effective generally from Term 3 2026, Semester 2 2026, and Hexamester 1 2027, with clauses 1.1–1.2 and 1.4 effective from 1 January 2027, and clauses 2.1–2.2 and 3.5–3.7 effective from 1 January 2028.
• draft-claim-unsw-chatgpt-edu-2026 — A UNSW announcement states that ChatGPT Edu access is being extended to Higher Degree Research candidates and casual academic staff, building on existing professional and academic staff access. It describes planned expansion of AI tools across the student community ahead of the 2027 academic year. Assessment permission still depends on the course instructions.
• draft-claim-unsw-information-governance-procedure — UNSW's Information Governance Policy applies to students, researchers, research trainees, employees and affiliates; individual procedures may have narrower scope. Its Use of Artificial Intelligence Systems or Tools Procedure requires users to maintain confidentiality of highly sensitive, sensitive, commercial and personal information throughout the data management life cycle. Before an AI system or tool is used, an AI self-assurance assessment must be conducted, except when using an approved, deployed enterprise-grade system such as Copilot.
• draft-claim-unsw-library-tdm-guidance — UNSW Library provides guidance for researchers and teaching academics using AI with copyrighted, licensed scholarly resources. The guide lists restrictions on automated scraping, reverse engineering and training AI systems on publisher datasets, and says publishers require attribution of AI use in submitted publications. It warns that publisher positions cannot be confirmed with100% accuracy, terms change, and researchers should check the relevant publisher or journal rules directly.