Guidelines on the use of AI

This guidance is intended to support graduate students and supervisors at UNB in the responsible, transparent, and scholarly use of Artificial Intelligence (AI) tools in graduate research. It complements:

Foundational principles

Across UNB policies and emerging sector-wide frameworks, four principles consistently govern acceptable AI use in graduate

    1. Students remain fully responsible for their work, regardless of the tools they use.
    2. Transparency matters more than prohibition: appropriate AI use is allowed when clearly disclosed and justified.
    3. Supervisors and students must align on use: expectations must be discussed early and revisited as research evolves.
    4. Ethics, privacy, and employment obligations apply: AI does not exempt students from existing rules.

These principles align with current graduate-sector practice that treats AI as a research support tool, not an author, collaborator, or decision-maker.

Student–supervisor and AI use disclosure checklists

Students and supervisors are expected to discuss AI explicitly at the beginning of the research relationship, using the Student–Supervisor Checklist and the AI Use Disclosure Worksheet as the primary mechanisms to:

  • Clarify whether and how AI tools may be used;
  • Identify tasks where AI use is permitted, restricted, or prohibited;
  • Establish expectations for documentation, disclosure, and verification;
  • Set boundaries related to ethics approval, data handling, and confidentiality.

AI expectations should be revisited at key milestones (proposal approval, ethics submission, thesis drafting).


1. Idea formation and early exploration

Common uses

  • Brainstorming research questions
  • Identifying potential gaps in literature
  • Exploring alternative theoretical framings

Guidance

  • AI may be used as a sparring partner, not as a source of authoritative knowledge.
  • All AI-suggested concepts must be validated against peer reviewed sources.
  • Students are encouraged to maintain informal AI reflection notes documenting how ideas evolved.

Supervisor discussion

  • How much ideation support is appropriate?
  • Must early research questions be developed without AI assistance?

2. Proposal development

Common uses

  • Structural outlining
  • Improving clarity, grammar, or flow

Restrictions

  • AI should not generate the core scholarly argument, research design, or rationale unless explicitly approved.
  • Some programs may prohibit AI drafting at this stage to assess independent scholarly voice.

Good practice

  • Include a brief AI transparency statement if AI was used.
  • Be prepared to explain and defend all proposal content without AI assistance.

3. Conducting research and analysis

Permissible uses (with supervisory endorsement)

  • Code generation or debugging (e.g., Python, R)
  • Supporting qualitative coding or thematic clustering
  • Summarizing researcher provided documents

Critical safeguards

  • Ethics approval is required before uploading sensitive or identifiable data into AI tools.
  • AI-assisted analysis must be described in the Methods section, including tool names and roles.
  • Outputs must be independently verified for accuracy, bias, and completeness.
  • You are ultimately responsible for the output generated. Recognize that hallucinated sources or phantom data generated by AI tools can lead to academic integrity and/or research misconduct concerns.

4. Thesis, dissertation and major research outputs

Accepted practices

  • Editing for clarity and language
  • Formatting assistance
  • Limited structural feedback

Mandatory disclosure

  • Students must include a Statement of AI Transparency facilitated by using the AI Use Disclosure Worksheet to state whether AI was used, which tools were used and, for what purposes.

The Statement of AI Transparency provides a standardized mechanism for this disclosure and is intended for inclusion in thesis front matter.


Prompt logging (recommended practice)

In addition to the required Statement of AI Transparency, students are encouraged (not required) to retain:

  • Key prompts;
  • AI-generated drafts or code;
  • Iterative refinements where AI meaningfully shaped output.

Prompt logs:

  • Are not part of thesis submission;
  • May be requested by supervisors or committees for clarification;
  • Support transparency in cases of authorship or integrity review.

As Students, graduate students often operate in dual roles, each with distinct AI implications.

Students are governed by:

In this role:

  • AI use must support learning and original scholarship;
  • Undisclosed or prohibited AI use may constitute academic misconduct.

Significant concerns are managed through:

  • Academic Incident Reporting processes and senate-based review

As Employees, when acting as paid research assistants (GSRA) or teaching assistants (GSTA):

Students must:

  • Confirm whether AI tools are permitted for employment duties;
  • Never upload confidential, proprietary, unpublished or personal data (e.g. student or participant data) into public AI systems;
  • Follow unit specific AI restrictions when applicable.

Significant concerns are managed through:

  • Research misconduct reporting and investigation processes;
  • Disciplinary processes detailed within the applicable collective agreement for the position held.