SciSpace Academy
Tutorial

Using AI Responsibly in Research: Risks, Rules and Best Practices

Dr. Faheem UllahEnglishResponsible AI

TL;DR

1-min read

AI tools now reach almost every part of research, from finding literature to writing and analysis, and they bring real risks to integrity and credibility. In this tutorial, Dr. Faheem Ullah, Assistant Professor and Cyber Security Program Director at the University of Adelaide, explains what responsible AI use looks like in practice for PhD students, postdocs, faculty and academic researchers who want to adopt AI in line with institutional and publishing expectations. The tutorial also gives practical guidance on using SciSpace responsibly across a research workflow, so that researchers benefit from AI without compromising originality or credibility.

Key takeaways

  • Hallucinations are outputs that sound plausible but are false, such as invented facts, misattributed findings or citations to papers that do not exist; every reference and claim should be checked against the original source.
  • Fabricated citations and over-reliance on AI are among the most common problems, and both can quietly undermine the originality of a piece of work.
  • Journals, reviewers and institutions increasingly expect authors to state how AI tools were used, especially for writing or analysis, and AI cannot be credited as an author.
  • Requirements vary, so check your target journal's policy and your institution's guidelines before submitting AI-assisted work.
  • The tutorial shares best practices for using AI in literature reviews, manuscript writing and synthesis, separating the tasks where AI helps from those that need human judgment.
  • Researchers should keep control of the research question, design, methods, evidence assessment, interpretation and argument, while AI can support searching, summarising and drafting.
Apply it

To put this into practice, run your draft through SciSpace's Reference Checker or Find, Add & Audit Citations agent and confirm each source before submission.

How to do it

  1. Define responsible use for your work

    Decide what responsible AI use means in your day-to-day research. Clear boundaries make using AI responsibly in research practical.

  2. Check for hallucinations and fabricated citations

    Verify every AI-generated claim and reference against the original source. This catches hallucinations and fabricated citations.

  3. Review journal and institution rules

    Check how your target journals, reviewers and institution view AI-assisted work. This keeps your work acceptable to them.

  4. Apply AI to reviews and writing carefully

    Use SciSpace's literature review and AI Writer for literature reviews, manuscript writing and synthesis. Follow best practices at each step.

  5. Keep human judgment in control

    Let AI add value where it helps, but make key decisions yourself. This avoids over-reliance on AI.

Watch the tutorial44 min

Using AI Responsibly in Research: Risks, Policies and Best PracticesPlay video: Using AI Responsibly in Research: Risks, Policies and Best Practices

Chapters

  1. Welcome and agenda
  2. Can ChatGPT write your research paper? The ethics line
  3. Where to use AI tools in research, from idea to review
  4. Why researchers use AI tools: speed and productivity
  5. Can one copied sentence get your paper rejected?
  6. What are the ethical risks of using AI in research?
  7. Does ChatGPT make up fake references?
  8. Is ChatGPT-written text plagiarism? Privacy and bias risks
  9. Should AI write your literature review or thesis?
  10. How to disclose AI use in a research paper
  11. AI disclosure statement template for papers and proposals
  12. Who is responsible for AI-generated mistakes?
  13. How to verify AI output and catch fake references
  14. Which research tasks to give AI vs do yourself
  15. Is it safe to upload research data to AI tools?
  16. AI detectors: what a 90% AI score means for your paper
  17. Do universities allow AI tools in research?
  18. How to acknowledge AI in methods, captions and journals
  19. What is ethical vs unethical AI use in research?
  20. Key takeaways and closing

Try this on SciSpace

Frequently asked questions

What are AI hallucinations in research?

Hallucinations are outputs that sound plausible but are false, such as invented facts, misattributed findings or citations to papers that do not exist. They are a serious risk in academic work. Prevent problems by verifying every reference and claim against the original source before using it in your writing.

Do I need to disclose AI use in my research paper?

Many journals and institutions now require authors to state how AI tools were used, especially for writing or analysis. Requirements differ, so check your target journal's policy and your institution's guidelines. AI tools cannot be credited as authors, and you remain responsible for all content.

Where should human judgment stay in control when using AI?

Researchers should decide the research question, design and methods, assess evidence quality, interpret results and form the argument. AI can help with searching, summarising and drafting, but conclusions and accountability belong to the researcher. Reviewing AI output critically is part of responsible use.

Related tutorials