AI is not a threat to research, but a powerful helper: BRUR VC
RANGPUR - At Begum Rokeya University, Rangpur (BRUR), the Vice-Chancellor affirmed a clear stance: "AI is never a threat to research; rather, it is playing a very effective role in improving the research quality and saving time."
He spoke at the opening of a two-day "Workshop on the Application of AI Tools in Academic Research," hosted by the Department of Computer Science and Engineering and organized by the Institutional Quality Assurance Cell (IQAC). The event was chaired by IQAC Director Prof Dr Md Tajul Islam, with Prof Dr Imran Mahmud (Head, Department of Science and Information Technology, Daffodil International University) delivering the keynote on using AI tools in academic research. Faculty from various departments joined, alongside Dr Md Abdur Rakib and Dr Md Sajib Mia.
"In fact, AI has brought about a radical change in the conventional framework of research and is gradually establishing itself as a powerful helper of academic research," the VC added.
What this means for scientists and scholars
Translation for day-to-day research: use AI where it trims friction, speeds iteration, and strengthens rigor.
- Literature triage: map fields, cluster themes, and draft reading lists before deep reading. Always verify sources.
- Method design: compare analytical approaches, generate checklists, and surface pitfalls you might miss under time pressure.
- Data work: assist with cleaning, feature ideas, and exploratory analysis; auto-generate code snippets with unit tests.
- Results review: probe alternative explanations, stress-test assumptions, and flag potential confounders.
- Writing and revision: structure sections, tighten prose, format references, and prepare lay summaries or abstracts to spec.
- Reproducibility: create notebooks with explicit steps, parameter logs, and data provenance notes.
Guardrails you can implement today
- Transparency: disclose where and how AI tools were used in the study and in manuscript preparation.
- Attribution: AI tools are not authors; authors remain accountable for all content and conclusions.
- Data protection: avoid pasting sensitive or proprietary data into tools that transmit to external servers.
- Bias checks: audit datasets and model outputs for skew; document mitigation steps.
- Verification: fact-check citations, recompute key figures, and cross-validate code generated by assistants.
- Local vs. cloud: prefer local or institution-vetted solutions for controlled datasets and regulated domains.
Department-level moves that pay off
- Run short, role-specific workshops for PIs, postdocs, and grad students focused on concrete workflows.
- Publish a one-page AI use policy: disclosure, data handling, and tool approval criteria.
- Maintain a vetted tool list with versioning and notes on acceptable use cases.
- Set up reproducibility templates (notebooks, reporting checklists) that integrate with AI assistants.
- Track time saved and error rates before/after adoption to quantify impact.
The VC underlined the urgency: "We are now living in a technology-dependent era, where there is no opportunity to lag behind. It is possible to make academic research more dynamic and standardized by acquiring a clear understanding and skills about AI." He emphasized increasing intelligent use of AI to raise research excellence.
Standards and further learning
- Practical resources on AI in academia: see Research.
- Ethics and disclosure guidance: see the Committee on Publication Ethics (COPE).
- Upskilling for researchers: curated options at Complete AI Training - courses by job.
Your membership also unlocks:
