PDF link to guidance: Guidelines for the Ethical and Responsible Use of GenAI in Grad Studies at UNBC (2025)

Purpose

To provide an ethical framework that supports the responsible, transparent, and equitable use of generative AI (GenAI) tools in graduate education at UNBC. This includes guidance across all stages of the research process - admission, problem framing, study design, data analysis, interpretation, and writing.

---

Guidelines

These guidelines are intended to help graduate students use AI in ways that uphold academic integrity and research ethics. They outline commonly accepted practices and highlight potential uses that may be inappropriate or unethical. Students are encouraged to consult their department and supervisors for discipline-specific guidance on the appropriate use of AI tools.

Audience

These guidelines represent the UNBC GenAI in Graduate Studies Research ethics baseline. Specific user guides will be created for grad students, staff, and faculty.

Foundational Principle: Ethics-of-Care

All AI use in research must prioritize an ethics-of-care approach that considers data ownership, community impact, research integrity, and Indigenous research protocols and data sovereignty. AI is to be used in ways that are transparent, responsible, and respectful of human subjects, communities, and research collaborators.

---

1 - UNBC Academic and Non-Academic Conduct Policy Note

UNBC has a student policy on Academic misconduct. The purpose of the policy is:

• Provide clarity, principles, standards and expectations for Students for both Academic and Non-Academic Conduct.
• Define Students’ responsibilities and rights as members of the University Community.
• Provide clarity and transparency for Students with respect to procedural and decision-making authority for Academic and Non-Academic Misconduct, investigations and Appeals.

Please see item 5.1.4 for relevance to these guidelines.

'5.1.4 Any action that violates the generally accepted standards of Academic Integrity is prohibited and deemed to be Academic Misconduct for the purposes of this Policy, including any act of dishonesty, falsification, misrepresentation, or deception in one’s academic work.' https://www.unbc.ca/sites/default/files/sections/policy/academicandnon-academicconductpolicy.pdf 

Failure to follow these guidelines may be considered academic misconduct. See also UNBC Graduate Calendar, general regulation III: Academic Conduct and Non Academic Conduct, https://www.unbc.ca/calendar. 

2 - Current practices of AI in Graduate Studies

a. Research design and proposal development

• Idea generation: Helps brainstorm research questions and align them with appropriate frameworks or methods.
• Literature scanning: Use AI tools (e.g., Elicit, Scite, ResearchRabbit) to find key papers, summarise insights, map trends, and streamline reviews.
• Question framing: Leverage AI to generate and refine research questions and position your contribution within existing literature.
• Writing support: Employ language models to draft or polish proposal sections (significance, methodology), with critical oversight and accurate citations.

b. Data collection and analysis

• Survey Design: Use AI to frame question formats, scales, and logic flows.
• Virtual Interviews & Focus Groups: Host sessions via Zoom’s AI Companion for real-time captions, automated transcription, meeting summaries, and sentiment insights.
• Statistical Analysis: Leverage AI-powered SPSS assistants or R packages to choose and interpret methods.
• Qualitative Coding: Use tools, such as Quirkos AI or NVivo AI, for initial coding and pattern detection.
• Visualization: Helps generate charts, tables, and figure captions for clarity.

c. Academic writing and publishing

• Drafting & Editing: Employ AI writing assistants for grammar, structure, and clarity.
• Multilingual support: Assists non-native English speakers with translation and expression.
• Productivity tools: Speeds up reference formatting and repetitive editing tasks.
• Citation Management: Use AI tools to organise sources, auto-generate citations, and enforce style consistency.
• Journal Selection: Try platforms (e.g., Elsevier Journal Finder) to match your manuscript with suitable
  outlets.

d. Dissemination and communication

• Presentation support: Prepares slide content, abstracts, and summary statements.
• Lay communication: Translates complex findings into accessible language for a broader audience.

e. Project management

• Task Automation: Automate transcription (e.g., Otter.ai), file conversions, scheduling, and other repetitive tasks.
• Planning Tools: Use AI-enhanced apps (Notion AI, Trello AI add-ons) for tracking tasks, milestones, dashboards, and workflows.
• Note: Some data analysis tools, such as SPSS and Atlas. Ti, NVivo, and R have built-in AI features.

3 - Risk-Based Tiered Guidelines

Minimal Risk Applications

• AI grammar/spell checks, paraphrasing non-sensitive public data.
• Permitted with supervisor awareness.

Moderate Risk Applications

• AI summarizing internal, anonymized or low-sensitivity data.
• Requires disclosure in thesis, coursework, or ethics applications.
• Supervisor approval is mandatory.

High Risk Applications

• AI processing of human subject data, Indigenous data, sensitive personal data.
• Supervisor approval is mandatory.
• Prohibited without explicit REB (Research Ethics Board) approval.
• Requires a data management plan and participant informed consent.

4 - Program and Discipline Autonomy 

• UNBC will establish a university-wide AI research ethics baseline.
• In addition to the UNBC ethics baseline, each Faculty/Program should develop discipline-specific guidelines articulating what is acceptable/unacceptable within their research contexts. If no guidelines are provided by the Faculty/Program, university AI guidelines will be followed.
• Discipline guidelines to be reviewed annually.

Examples:

• Computer Science: Synthetic data permitted.
• Social Sciences: AI-generated data prohibited in case studies.

5 - Education-First, How-Not-No Model

• AI literacy workshops for all graduate students and faculty.
• Clear guidelines on AI limitations, citation standards, and responsible use.
• Ongoing, discipline-specific workshops.
• Supervisors to address AI use expectations with all students.

6 - Process-Over-Product Assessment

• Emphasize process transparency over AI detection.
• Require AI use disclosures in these proposals and coursework.
• Encourage version logs, tracked changes, and research process reflection sections.
• Integrate ethical reasoning and AI-related decisions into oral defenses.

7 - Consent, Disclosure, and Data Governance

1. AI use of inhuman subject data requires explicit participant informed consent.
2. Uploading sensitive data to unsanctioned AI tools is prohibited.
3. Mandatory data security protocols are required for AI-generated or processed content.

Consent form addendum:

“This research involves the use of artificial intelligence tools for data analysis/text generation, including specific
uses __________________.

“Your data will/will not be processed through external AI systems. You may choose to opt out.”

• REB applications must address AI use explicitly in intentions.

Researcher Responsibility:

Researchers are fully accountable for the content and interpretation of any output generated by generative AI tools. These tools are probabilistic and should not be treated as authoritative sources. Researchers must critically evaluate, verify, and, where necessary, correct any AI-generated material before including it in academic work. Any errors that result from the use of such content—whether attributed or not—remain the responsibility of the researcher.

8 - (3) Decision Trees for Ethical Use of AI:

1.  Is AI being used in research?
   1. No → Standard academic conduct applies.
   2. Yes.
      1. Human subject data involved?
         1. No → Program-specific guidelines + disclosure.
         2. Yes.
            1. Has REB approved AI use?
               1. Yes → Proceed with safeguards.
               2. No → Stop. Submit revised ethics.
2. Is AI used in writing the thesis/dissertation/capstone/project?
   1. No → Standard academic conduct applies.
   2. Yes - UNBC Academic Conduct and Non-Academic Conduct Policy applies.
      1. Used AI to find literature or summarize papers/articles?
         1. No → Standard academic conduct applies.
         2. Yes → Program-specific guidelines + disclosure.
            1. Used AI to write parts of the thesis/dissertation?
               1. No → Standard academic conduct applies.
               2. Yes → Program-specific guidelines + disclosure.
                  1. Used AI to create the list of references?
                     1. No → Standard academic conduct applies.
                     2. Yes → Program-specific guidelines + disclosure.
3. AI use risk level?
   1. Minimal (grammar check).
      1. →Permitted with disclosure.
   2. Moderate (internal summaries.
      1. →Disclosure + supervisor sign-off.
   3. High (raw human data, Indigenous data).
      1. →Disclosure + supervisor sign-off + REB review + consent.

*Individual consent is required for participants in human subjects research. In addition, the Indigenous Nation or Organization must provide explicit, prior, and collective consent before their data or knowledge is used in AI systems. For further Guidance on Indigenous Knowledge Sovereignty see First Nations principles of OCAP®