The University of Manchester

Under Academic Malpractice Procedure DocID 639 §4.4.2, where indicators raise concerns of commissioned work without direct evidence, or available information does not satisfy the balance of probabilities, a viva voce can be arranged in agreement with the School Officer to demonstrate authorship; it does not contribute to the mark. If held, it should not normally take place more than 30 working days after the assessment feedback deadline (§4.4.3.1); the student should receive at least five working days notice explaining its nature, purpose and potential outcomes (§4.4.3.2). Accompaniment by a person of the student’s choice is for support only without participation (§4.4.3.4), and reasonable adjustments must be made (§4.4.3.5). The viva itself cannot impose a penalty (§4.4.4.2); case handling follows the applicable AUO/SDP/UDP procedures (§5.1–5.5).

Under Academic Malpractice Procedure DocID 639 §3.1.3.6 and Senate Policy DocID 79331 §3.0.2, using content generated by Artificial Intelligence without acknowledgement is explicitly designated as an indicator of academic malpractice and plagiarism.

Under Academic Malpractice Procedure DocID 639 §4.2.2, the University explicitly does not use software that claims to detect Generative AI; Turnitin is used solely for text-matching similarity detection and cannot confirm whether malpractice occurred without marker academic judgement.

Under Senate Policy DocID 79331 §2.1.3, the designated AI category for every assessment must be clearly communicated to students in the Course Unit Specification and within the assessment brief on the virtual learning environment (Canvas).

Under Senate Policy DocID 79331 §2.1.1(d), assessments designated as Category D (AI-Integrated) require students to use AI tools to achieve the unit's intended learning outcomes, with the assessment brief detailing the tools, permitted uses, reflection, and logging requirements.

Under Senate Policy DocID 79331 §2.1.1(b), assessments designated as Category B (AI Minimal) permit using AI tools only in relation to presentation—such as to improve fluency, grammar, or spelling, or to assist with formatting—and strictly forbid using AI tools to generate ideas, arguments, conclusions, or other substantive content.

Under Senate Policy DocID 79331 §2.1.1(c), assessments designated as Category C (AI Permitted) allow students to use AI tools for wider tasks, but require assessment briefs to state which tools and outputs are allowed, mandate that submitted text must be authentic student work, and require that a non-AI pathway exists so students are not disadvantaged if they choose not to use AI.

Under Senate Policy DocID 79331 §2.1.1(a), assessments designated as Category A (AI Prohibited) completely forbid using AI tools to assist in the preparation or writing of the submission, including for generating content or for proofreading / editing text, and the assessment brief must explain why this prohibition is in place.

Under Senate Policy DocID 79331 §2.1.4 and §2.1.5, course unit leads must define AI categories and logging for assessed group work, and students must seek prior agreement from all group members before recording or transcribing group discussions with AI tools.

Under Senate Policy DocID 79331 §2.1.2, students may choose not to use AI for assessment, unless the assessment has been designated as Category D (AI-Integrated); students in Categories A, B, and C who choose not to use AI will not be disadvantaged.

Under Senate Policy DocID 79331 §2.2.2, students are explicitly permitted to upload session recordings into the University-provided Microsoft 365 Copilot service to ask questions or create study summaries.

Under Senate Policy DocID 79331 §2.2.1(c), students must not upload any University of Manchester-owned copyright materials—such as lecture recordings, PowerPoint slides, handouts, and examination question papers—to external commercial AI tools.

Under Senate Policy DocID 79331 §2.3.1, staff members who use AI in Teaching and Learning activities must adhere to the University's AI Guidelines, and only the University-provided Microsoft 365 Copilot service may be used for student-identifiable or confidential data.

Under StaffNet Research Guidelines, corresponding authors carry ultimate responsibility for the accuracy, balance, and credibility of work submitted for publication, must comply with publisher guidelines, and must clearly declare any significant AI assistance.

Under StaffNet Research Guidelines, researchers reviewing papers for publication or funding applications must preserve peer review confidentiality and must not upload unpublished manuscripts, grant proposals, or data into external AI services.

Under StaffNet Teaching and Learning Guidelines, where AI use is encouraged or required, the University must ensure all students have equitable access at no additional cost or provide an alternative mechanism that avoids requiring students to provide personal data to third parties.