C.7 — AI Usage Policy

Document Control Information

Policy Code

C.7

Standard

C. Education Program

Status

published

Version

5

Process Owner

Assoiate Dean Academic Affairs(ADAA)

Prepared By

ADAA

Reviewed By

Academic Council

BOT Approval

approved

First Issue Date

06 Nov 2024

Last Review Date

15 Oct 2026

Relevant To

DMU Faculty,Students

Used With

—

Modification History

Purpose and Scope

• AI-powered educational platforms and personalized learning systems.
• AI-driven research tools used for data analysis, simulations, and predictive modeling.
• Generative AI tools like ChatGPT for creating content, assignments, and research projects.
• AI tools used in clinical simulations, patient care scenarios, and decision-making processes.

Abbreviations and Definitions

• AI (Artificial Intelligence): Computer systems capable of performing tasks that typically require human intelligence, such as reasoning, learning, and decision-making.

• Generative AI: AI systems capable of producing new text, images, audio, code, or other content in response to a prompt.

• Large Language Model (LLM): A type of Generative AI trained on large volumes of text to understand and generate human-like language.

• Human Oversight: The requirement that a qualified person reviews, verifies, and takes responsibility for AI-assisted outputs before they are relied upon.

• AI Hallucination: Content generated by an AI system that is factually incorrect, fabricated, or unsupported by evidence, but presented as if accurate.

• Algorithmic Bias: Systematic errors or unfair outcomes produced by an AI system that disadvantage particular individuals or groups.

• AI Incident: Any event in which an AI tool causes, or could reasonably cause, harm, a data breach, a safety risk, or a significant policy violation.

• Sensitive Data: Patient identifiers, clinical records, student records, staff records, unpublished research data, examination materials, or other confidential institutional information.

• AI System Inventory: An institutional register listing all AI tools approved for use, their purpose, and their risk classification.

• Agentic AI: AI systems capable of autonomously carrying out multi-step tasks or actions on a user's behalf with limited direct human input.

Policy Statement

All users of AI tools and Generative AI technologies must:

• Adhere to ethical guidelines: Ensure the responsible and ethical use of AI in accordance with institutional and UAE regulations.
• Maintain academic integrity: Avoid using AI tools to engage in plagiarism, academic dishonesty, or other unauthorized practices.
• Ensure accuracy: Verify and cross-check outputs from AI systems to ensure reliability and credibility.
• Protect data privacy: Follow data protection protocols when using AI systems that involve sensitive information, particularly in clinical and research contexts.
• Promote inclusivity: Utilize AI tools in a way that fosters equitable access and avoids biases or discrimination.

• Report misuse: Notify relevant authorities if AI tools are used inappropriately or in violation of this policy​​

Policy Statement — Continued

DMU’s AI usage policy is built upon the following ethical principles:

·         Transparency: All AI use in academic work must be disclosed, and students/faculty must clearly indicate when AI tools have been utilized in research, assignments, clinical simulations, or content creation.

·         Accountability: Human oversight is essential for all AI-driven decisions. AI tools must complement, not replace, human judgment in both educational and clinical contexts. Faculty members are responsible for verifying the outputs of AI tools used in assessments and simulations.

·         Fairness and Non-discrimination: AI tools must be regularly audited to ensure they do not introduce or perpetuate biases. This applies to AI systems used in student assessments, research, and clinical decision-making.

·         Data Privacy and Security: The use of AI tools must adhere to UAE’s Federal Law No. 2 of 2019, GDPR, and HIPAA to protect sensitive patient and academic data. AI systems must not compromise patient confidentiality or expose private information.

·         Beneficence and Non-maleficence: AI tools must enhance educational outcomes and research while ensuring that no harm is caused through misinformation, bias, or over-reliance on AI.

·         Human Oversight: A qualified human must remain responsible for, and able to override, any decision informed by AI.

·         Integrity: AI must be used honestly and disclosed appropriately in all academic, research, and clinical work.

·         Equity: AI tools and their benefits must be made accessible to all students, faculty, and staff regardless of background or ability.

·         Verification: All AI-generated outputs must be checked against reliable sources before being used or relied upon.

·         Safety: AI tools must not be used in ways that could endanger patients, students, staff, or the integrity of the institution.

Procedures & Guidelines

Educational Content Creation: Faculty members may utilize AI tools to enhance course materials, lectures, and case studies. However:

•AI-generated content must be reviewed for accuracy and academic rigor.

•AI should supplement, not replace, critical academic content developed by faculty.

1.       Faculty Responsibilities (proposed addition): Faculty may use AI tools to support the development of learning outcomes, rubrics, quizzes, case scenarios, curriculum mapping, feedback drafting, reports, and meeting minutes.

Faculty must not:

• Allow AI to mark or grade student work independently, without faculty review.

• Allow AI outputs to determine student progression or academic standing.

• Upload confidential assessments, question banks, or student records into public AI tools.

2.       Student Use of Generative AI (e.g., ChatGPT and Similar Tools): Students are permitted to use Generative AI tools for research, content generation, and studying under the following conditions:

·         AI-generated content must be clearly cited in all assignments and projects.

·         Students must disclose their use of AI in submitted work to maintain academic integrity.

·         Misuse of AI (e.g., submitting AI-generated essays, reflections without acknowledgment) will be treated as academic misconduct, with penalties in accordance with DMU’s academic policies.

·         Students may use AI tools for concept clarification, generating practice questions, brainstorming ideas, improving grammar and language, summarizing personal notes, creating study plans, and supporting literature searches where permitted. Students remain responsible for the accuracy, originality, and integrity of all submitted work.

3.       Examinations and Assessments:

AI use in assessments shall be determined by the course instructor and clearly stated in the course syllabus or assessment instructions.

·         Category A – AI Prohibited: No AI use is permitted except spelling, grammar, or accessibility tools. This category applies by default to examinations, OSCEs, viva examinations, in-class tests, and clinical competency assessments.

·         Category B – AI Permitted with Declaration: AI may be used for specified purposes, provided the student declares how AI was used.

·         Category C – AI Integrated: AI use forms part of the assessment task, and students are required to demonstrate critical evaluation and verification of AI-generated outputs.

AI Declaration:

Where AI is used in assignments, projects, reports, theses, dissertations, or scholarly work, users should include the following declaration:

"I used [AI Tool Name] for [purpose]. I reviewed, verified, and take full responsibility for the final submitted work."

Assessment Redesign (proposed addition): Where feasible, assessment design should evolve to reduce reliance on AI-detection tools alone by increasing the use of assessment formats that are harder to complete solely with AI, such as viva voce examinations, OSCEs, portfolios, oral defence, reflective writing, and other authentic, performance-based assessment.

Procedures & Guidelines — Continued

Use of AI in Research

AI-Assisted Research

·         Researchers are encouraged to use AI tools for tasks such as data analysis, predictive modelling, and simulations.

·         All AI tools and methodologies employed must be disclosed in research publications and reports.

Peer Review (proposed addition): Faculty and reviewers must not upload manuscripts, grant applications, or confidential reviewer reports into public AI tools.

AI Contributions

·         Research publications must acknowledge the role of AI tools, including their specific contributions.

·         Limitations of the AI tools should be transparently stated to avoid over-reliance on AI-generated data.

Research Integrity and Authorship (proposed addition):

·         AI tools cannot be listed as an author on any publication.

·         AI must not be used to fabricate data, references, citations, or research participants.

·         AI cannot replace scientific or scholarly judgement in the design, analysis, or interpretation of research.

Intellectual Property (proposed addition): Ownership and copyright of AI-assisted institutional material must be clarified in accordance with DMU policy and applicable UAE law before publication or distribution.

AI in Clinical Simulations

Clinical Education

·         AI-generated simulations can be used to enhance student training by providing real-world decision-making scenarios.

·         Faculty members must review all AI-generated clinical scenarios for accuracy and alignment with real-world medical practices.

Explicit Clinical AI Limitations (proposed addition): AI must not be used to diagnose patients, prescribe treatment, or replace clinical judgement. AI-generated clinical content is for training and educational purposes only and must be verified by a qualified clinician.

AI-Generated Multimedia (proposed addition): The use of deepfakes, AI-generated images, or manipulated clinical images is prohibited except where explicitly approved for educational simulation purposes and clearly labelled as AI-generated.

Patient Safety Simulations

AI tools used in patient care simulations must be carefully monitored to ensure they do not generate harmful outcomes.

AI Governance

Dubai Medical University (DMU) has established an AI and Smart Education Unit to oversee the ethical and responsible use of AI tools across the institution. This unit collaborates with representatives as advisory members from the Academic Affairs Departments and Program heads of all colleges to ensure robust governance of AI systems and tools.

AI Governance Committee (proposed addition): DMU should establish an AI Governance Committee, with representation from Academic Affairs, Research, QAIE, IT, Legal, IRB, Faculty Development, Assessment, and Students, to provide broader institutional oversight of AI adoption and risk.

AI System Inventory (proposed addition): DMU should maintain an institutional register of approved AI tools, documenting their purpose, data-handling practices, and risk classification.

Approved AI Tools (proposed addition): Only institutionally approved AI tools should be used for institutional business, including teaching, assessment, research, and administrative work.

AI Procurement (proposed addition): Before purchasing or adopting a new AI system, DMU should conduct a security review, legal review, and vendor assessment.

Key responsibilities of the AI and Smart Education Unit include:

·      Policy Development: Developing and updating AI policies to align with emerging technological advancements and ethical standards.

·      Usage Monitoring: Supervising the application of AI tools by faculty, students, and researchers to ensure compliance with institutional policies and ethical principles.

·      Ethical Oversight: Conducting regular evaluations to identify and address biases, inaccuracies, or ethical concerns in AI systems.

·      Training and Support: Offering guidance and training to faculty and students for the responsible and effective use of AI in education, research, and clinical practice.

AI Literacy and Training (proposed addition): DMU should provide mandatory AI literacy training for students, faculty, and researchers, covering effective and responsible prompting, ethics, AI hallucinations, algorithmic bias, and confidentiality.

Monitoring and Audits

To ensure the ethical and effective use of AI, the AI and Smart Education Unit will implement the following measures:

Annual AI Reviews

Conduct comprehensive yearly evaluations of AI tools used in educational, research, and clinical settings to maintain alignment with ethical standards and institutional goals.

Bias Testing:

·         Regularly test AI systems, particularly those used in student assessments, clinical simulations, and research, to detect and mitigate biases.

·         AI-detection tools may be used to support academic integrity investigations; however, AI-detection results shall not be used as the sole basis for determining academic misconduct and must be considered alongside other evidence.

Student Wellbeing (proposed addition): Students must not be penalized solely on the basis of an AI-detection tool result where that result produces a false positive; findings must always be corroborated with other evidence.

Data Privacy and Security: DMU will enforce strict data protection protocols to ensure compliance with GDPR, HIPAA, and UAE Federal Law No. 2 of 2019. Users must not upload patient identifiers, clinical records, student records, staff records, unpublished research data, examination materials, or confidential institutional documents into public AI platforms unless specifically approved by the University.

Measures include:

•Data encryption to protect sensitive information.

•Controlled access to AI systems.

•Regular security reviews to prevent breaches and unauthorized use.

Additional Data Protection Requirements (proposed addition)

• Cross-border data transfer involving AI tools must comply with UAE data protection requirements and be reviewed before use.

• Vendors providing AI systems remain responsible for the security and appropriate handling of any institutional data processed by their tools.

• Data residency requirements must be confirmed for any AI system that stores or processes DMU data.

• AI procurement decisions must include a data protection assessment.

• Only vendors approved through DMU's procurement and legal review process may be used for AI systems handling institutional data.

Violations and Enforcement

Academic Misconduct: Misuse of AI tools, including Generative AI systems like ChatGPT, that compromises academic integrity will be treated as academic misconduct, with penalties as outlined in DMU’s academic policy.

Research Misconduct: Researchers who fail to disclose the use of AI tools or misuse AI in their projects will face disciplinary actions according to DMU’s research integrity guidelines.

Examples of AI- Related Academic Misconduct

The following may constitute academic misconduct:

·         Submitting AI-generated work without disclosure where required.

·         Using AI in examinations or assessments where prohibited.

·         Fabricating references, citations, clinical cases, patient information, or research data using AI.

·         Uploading another person's work into an AI system without permission.

·         Using AI to impersonate another individual.

·         Using AI to conceal plagiarism.

Progressive Discipline Framework (proposed addition): To support consistent enforcement, violations should be classified using a graded discipline matrix:

• Minor violations (e.g., undisclosed use of AI for low-stakes tasks): verbal or written warning and educational follow-up.

• Moderate violations (e.g., unauthorized AI use in an assessment): formal academic misconduct proceedings and grade penalties as appropriate.

• Major violations (e.g., fabrication of data or impersonation using AI): referral to the relevant disciplinary committee, with potential suspension or dismissal.

• Research misconduct involving AI: handled under DMU's research integrity procedures, which may include retraction and reporting to funders or journals.

• Data breaches involving AI tools: immediate escalation to IT/Legal and applicable regulatory notification.

Responsibilities for Application

• Academic affairs department of respective colleges
• AI and Smart Education Unit