Skip to main content
Legal

Responsible AI and Learning Policy

Last updated: 8 July 2026

AI should strengthen learning, not replace it

NuvStem supports responsible use of modern tools, including AI.

AI can be useful for explanation, feedback, practice, review, and exploration. It becomes harmful to learning when it is used to conceal authorship, fabricate evidence, bypass academic rules, or produce work the user cannot understand or defend.

Our principles

Learning comes first

AI should help a student understand a concept, test reasoning, find gaps, compare approaches, or improve a process.

The goal is not merely to produce output. The goal is to build understanding.

The user remains responsible

AI output can be wrong, incomplete, biased, insecure, outdated, or misleading.

Users must verify:

  • calculations;
  • code behaviour;
  • citations;
  • technical claims;
  • experimental conclusions;
  • safety assumptions; and
  • any claim used in academic, professional, or regulated work.

Rules and disclosure matter

Courses, schools, journals, employers, competitions, and professional bodies may impose different rules on AI assistance.

Users must follow the rules that apply to their work and disclose AI assistance where required.

No fabricated evidence

AI must not be used to invent:

  • sources;
  • citations;
  • quotations;
  • data;
  • experiments;
  • interviews;
  • survey responses;
  • calculations presented as measured results; or
  • professional experience.

Appropriate uses

Subject to the rules that apply to the user, appropriate uses may include:

  • asking for another explanation of a concept;
  • generating practice questions;
  • brainstorming an outline that the student will research and write;
  • checking a student's own draft for unclear reasoning;
  • reviewing code for potential bugs;
  • explaining compiler or runtime errors;
  • suggesting test cases;
  • comparing technical approaches;
  • identifying missing validation steps;
  • recommending relevant resources; and
  • helping a tutor organize feedback.

Uses NuvStem will not knowingly support

NuvStem will not knowingly support:

  • generating prohibited graded work for submission as a user's own;
  • producing answers the user intends to misrepresent;
  • fabricating data, sources, citations, experiments, or results;
  • using AI to impersonate a student;
  • bypassing proctoring or access controls; or
  • violating the academic-integrity rules that apply to the user.

How NuvStem currently uses AI

NuvStem may use locally operated AI models for limited purposes such as:

  • code review or code analysis;
  • internal tutor and support workflows; and
  • mentor or resource recommendations.

Under the current policy:

  • User Content and conversations are not used to train NuvStem models or third-party models unless the user separately gives express, informed consent;
  • NuvStem does not currently send User Content to third-party generative AI providers for these AI-assisted functions; and
  • AI-assisted recommendations remain subject to human judgment and do not guarantee suitability or outcomes.

For details about personal information and AI-assisted processing, see the Privacy Policy.