AI Risks and Limitations

Use AI transparently and build on academic sources to keep your work authentic.

Misusing AI could put your academic integrity at risk — for example, submitting AI-generated work without acknowledgement may be treated as academic misconduct. See our page on Rules and Responsibilities for full details.

Cross-check AI outputs with trusted academic databases to strengthen your arguments.

AI outputs may:

• Contain inaccurate information — AI tools frequently get facts wrong.
• Be outdated — Some AI tools may be trained on data that is months or years old.
• Lack understanding — AI does not “know” meaning or context, so outputs can be irrelevant or nonsensical.
• Hallucinate — AI may generate content or references that look convincing but are entirely false. Always verify them in LibrarySearch.

Safe action: Always check AI results against trusted academic sources, such as LibrarySearch or subject databases. As a general rule, do not cite AI output directly — instead, return to and reference the original sources. Only in specific cases (when required by your module or assessment) should AI outputs be cited, and in those cases follow our Referencing AI guide.

For detailed guidance and examples of how to spot and deal with hallucinations and fabricated references, see our page on Evaluating AI-Generated Content.

Apply critical thinking and seek diverse perspectives when using AI.

AI outputs may:

• Reproduce biases and stereotypes, including Western-centric perspectives.
• Perform unevenly, doing better in widely published subjects and worse in niche or specialist fields.
• Promote unfair, extreme, or unbalanced views.

Safe action: Apply critical thinking, look for diverse perspectives, and compare with peer-reviewed sources.

For more detail on spotting and addressing bias in sources (including AI outputs), see our guide on Recognising and Addressing Bias.

Use AI to support your learning, but make sure you continue to develop the skills your programme is designed to teach.

Over-reliance on AI may:

• Weaken essential abilities such as writing, coding, design, and critical thinking.
• Encourage shortcuts — relying on AI summaries instead of engaging with the full reading/listening to it limits deep understanding.
• Limit your future — if you skip building key skills now, it may make honours projects, postgraduate study, or employment more challenging.

Safe action: Use AI when permitted (or if specifically directed to in your assessment brief), but make sure you still practice and develop the skills your programme is designed to teach. Remember that systematic searching and critical reading are essential for dissertations, literature reviews, and postgraduate work — AI cannot replace these academic processes.

See also: Using AI for research for guidance on combining AI exploration with proper academic sources, AI for Writing for support in developing your own writing skills, and Using AI to Understand Academic Papers for help with reading and deep understanding.

Be careful what you share with AI tools to keep your personal data safe.

Privacy risks include:

• Anything you type may be stored or reused by the provider.
• Some AI tools may use your inputs to train future models.

Safe action: Use Abertay-supported tools where possible, keep extracts short, and never share personal or sensitive information.

For more detail on protecting your information, see our page on Student data privacy with AI.

Respect copyright and take responsibility for your work, even when AI supports you.

Ethical risks include:

• Copyright — AI outputs may include copyrighted material without permission. For more detail, see our page on Copyright for Students.
• Accountability — You remain responsible for the work you submit, even if an AI tool shaped it.

Safe action: Use AI outputs responsibly and never present them as entirely your own. Always check against reliable academic sources, and make sure your final work reflects your own ideas. If you do use AI (and your assessment brief permits it), ensure this is properly acknowledged.

Be aware that AI tools require significant computing power, which has energy and sustainability implications.

Environmental risks include:

• Increased energy use — Large AI systems can consume high amounts of electricity.
• Sustainability concerns — Widespread AI use contributes to a growing carbon footprint.

Safe action: Use AI thoughtfully and only when it adds real value to your learning, to help minimise unnecessary environmental impact.