10th Annual National Training Workshop
Central Library, University of Peshawar
Baragali Summer Campus, July 3–5, 2026
Artificial Intelligence in Higher
Education Research
Tools, Techniques, Ethics and Practical Workflows
Lecture Notes
Prepared for teachers, researchers, librarians, administrators and postgraduate
scholars
Dr. Shujaat Ali
Ask better Find quality Read and
Write and verify
questions literature organize
AI in Higher Education Research Dr. Shujaat Ali
Contents
How to Use These Notes 2
1 Workshop Learning Outcomes 2
2 Simple Introduction: What is AI in Research? 2
2.1 What AI Can and Cannot Do . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3 The Complete AI Research Workflow 3
4 Tool Map for Participants 3
5 Prompting: How to Talk to AI 4
6 ChatGPT and Claude AI 4
6.1 Practical Uses in a Workshop Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
6.2 Example: From Vague Topic to Research Problem . . . . . . . . . . . . . . . . . . . . . 5
6.3 Using ChatGPT and Claude Together . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
7 ResearchRabbit: Seeing the Literature Network 5
7.1 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
7.2 Activity: Build a Literature Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
8 Semantic Scholar: Searching Scholarly Literature 6
8.1 How to Search Better . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
9 NotebookLM: Turning Documents into a Knowledge Base 6
9.1 Three-Part NotebookLM Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
9.2 NotebookLM Workflow for Teachers and Librarians . . . . . . . . . . . . . . . . . . . . . 7
10 Consensus: Evidence-Based Answers 7
10.1 Good Questions for Consensus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
10.2 Consensus with ChatGPT and Claude . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
11 STORM: Topic Exploration and Knowledge Curation 8
11.1 When STORM is Useful . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
12 Jenni AI and Paperpal: Academic Writing Support 8
12.1 Where They Fit in the Writing Process . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
13 AI for Librarians 9
13.1 AI-Supported Library Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
14 AI for Teachers and Supervisors 9
14.1 Teaching Prompts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
15 AI for Administrators and Academic Managers 10
15.1 Administrative Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
16 Responsible AI Use: VERIFY Model 10
17 Academic Integrity and Ethics 10
17.1 Core Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1
AI in Higher Education Research Dr. Shujaat Ali
18 Prompt Library for Participants 11
18.1 Topic Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
18.2 Literature Review Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
18.3 Research Gap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
18.4 Methodology Explanation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
18.5 Manuscript Editing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
19 Hands-On Workshop Activities 11
20 Assessment: Quick Reflection Questions 12
21 Model Institutional Policy Points 12
22 Conclusion 12
Selected Sources for Tool Descriptions 13
2
AI in Higher Education Research Dr. Shujaat Ali
How to Use These Notes
These lecture notes are written in simple language for participants from different backgrounds. Some
participants may be teachers, some may be librarians, some may be research supervisors, and some may
be administrators. The aim is not to make everyone a computer scientist. The aim is to help everyone
use AI safely, practically, and responsibly in higher education research.
Main message of the workshop
AI should be used as a research assistant, not as a replacement for the researcher. AI can help
in searching, reading, summarizing, organizing, drafting and editing, but the human researcher
remains responsible for accuracy, originality, ethics and final judgment.
1 Workshop Learning Outcomes
By the end of this workshop, participants should be able to:
1. Explain the role of AI in higher education research using simple terms.
2. Select the right AI tool for a specific research task.
3. Use ChatGPT and Claude for brainstorming, explanation, planning, summarization and academic
support.
4. Use ResearchRabbit and Semantic Scholar for literature discovery and citation exploration.
5. Use NotebookLM to work with uploaded source documents in a controlled and traceable way.
6. Use Consensus, STORM, Jenni AI and Paperpal in appropriate stages of the research workflow.
7. Write safe and useful prompts for research, teaching and library services.
8. Apply ethical checks before using AI-generated material in academic work.
2 Simple Introduction: What is AI in Research?
Artificial Intelligence, or AI, means computer systems that can perform tasks that usually require
human intelligence. In research, AI can help with tasks such as finding papers, summarizing long
documents, explaining difficult concepts, improving writing, generating research questions, preparing
teaching material, and checking the logic of a draft.
2.1 What AI Can and Cannot Do
Research Stage AI Can Help With Human Must Still Do
Topic selection Suggest directions and refine Decide novelty, feasibility and
scope local relevance
Literature review Find, summarize and compare Verify sources, read key papers
papers and judge quality
Methodology Explain methods and draft Choose valid design and justify
workflow decisions
Writing Improve clarity, structure and Ensure originality, accuracy and
language academic voice
Publication Check formatting and cover Follow journal policy and ethical
letters rules
3
AI in Higher Education Research Dr. Shujaat Ali
Important rule
Never submit AI-generated content without reading, checking and rewriting it in your own academic
understanding. AI can make confident mistakes, invent references, misunderstand local context or
overstate results.
3 The Complete AI Research Workflow
The following workflow shows how different AI tools can support research from topic selection to final
manuscript preparation.
1. Topic and 2. Literature 3. Reading
research problem discovery and notes
ChatGPT, Semantic Scholar, NotebookLM,
Claude, STORM ResearchRabbit ChatGPT, Claude
revise
again
6. Editing and 5. Drafting 4. Synthe-
verification and structure sis and gap
Paperpal, hu- Jenni AI, Chat- Consensus,
man review GPT, Claude Claude, ChatGPT
Figure 1: A simple AI-supported research workflow.
4 Tool Map for Participants
Each tool has a different role. A good researcher does not use one tool for everything. The following
table gives a practical map.
Tool Best Use Simple Example Main Caution
ChatGPT Explaining ideas, planning re- Ask it to explain a the- Verify facts and cita-
search, drafting prompts, sim- ory to undergraduate stu- tions
plifying text, data interpreta- dents
tion support
Claude AI Long-document reasoning, Ask it to compare three Verify sources and
structured writing, careful theories in a table claims
explanation and feedback
ResearchRabbit Literature mapping, citation Start with one seed pa- Do not treat every re-
networks, related papers, au- per and discover related lated paper as high
thor connections clusters quality
Semantic Scholar Academic search, paper dis- Search latest papers on Read the actual pa-
covery, citations, related work AI in libraries per before citing
NotebookLM Work with uploaded docu- Upload workshop PDFs Answers depend on
ments, generate summaries, and ask for key themes uploaded sources
study guides and source-
grounded answers
Consensus Evidence-based answers from Ask whether an interven- Check paper quality
research literature tion is supported by stud- and context
ies
STORM Broad topic exploration and Generate a Wikipedia- Use as a starting
outline generation like overview of AI policy point, not final evi-
in HE dence
4
AI in Higher Education Research Dr. Shujaat Ali
Tool Best Use Simple Example Main Caution
Jenni AI Academic drafting, citation- Draft a paragraph from Avoid generic writing
supported writing workflow notes and cited sources and check citations
Paperpal Academic editing, grammar, Improve a manuscript Editing cannot fix
clarity, submission readiness section before submission weak research design
5 Prompting: How to Talk to AI
A prompt is the instruction we give to an AI tool. Better prompts usually produce better answers. A
weak prompt is vague. A strong prompt gives role, task, context, output format, limits and verification
requirements.
Role Task Context
Who should What should What is the
AI act as? it do? background?
Format Level Checks
Table, bullets, Beginner, Ask for limits
outline expert, mixed and cautions
Figure 2: The six parts of a strong academic prompt.
General research prompt template
Act as an academic research assistant. I am working on [topic] for [audience or level]. Help me
[specific task]. Use simple language. Provide output in [table / outline / steps]. Include possible
limitations, ethical issues and verification steps. Do not invent citations. If evidence is needed, tell
me what sources I should check.
Prompt for simplifying a difficult concept
Explain [concept] to participants from mixed backgrounds, including teachers, librarians and
research students. Use simple examples from higher education. Give one short definition, one
analogy, one practical use, and one caution.
6 ChatGPT and Claude AI
ChatGPT and Claude are general-purpose AI assistants. In higher education research, they are useful for
planning, explanation, brainstorming, writing support, summarization, teaching material and reflection.
They are not databases. They should not be trusted blindly for exact facts or references.
Official educational descriptions present these tools as assistants for students, faculty, researchers and
academic operations, with emphasis on responsible use, security and learning support. ChatGPT Edu
is described by OpenAI as a university-focused version of ChatGPT for students, faculty, researchers
and campus operations. Anthropic describes Claude for Education as supporting AI-enabled approaches
across teaching, learning and administration.1
6.1 Practical Uses in a Workshop Setting
• Convert a vague topic into a focused research problem.
• Generate research questions and objectives.
1
OpenAI, Introducing ChatGPT Edu, 2024; Anthropic, Introducing Claude for Education, 2025.
5
AI in Higher Education Research Dr. Shujaat Ali
• Explain a theory in simple terms.
• Create a seminar plan, lesson plan or library training session.
• Summarize a document, but only after providing the document.
• Improve writing clarity without changing the meaning.
• Create rubrics, checklists and feedback templates.
6.2 Example: From Vague Topic to Research Problem
Prompt
Act as a research supervisor. My broad topic is: AI use by university librarians. Help me convert it
into three focused research problems. For each problem, provide a possible title, objective, research
question, methodology and expected contribution. Use simple language.
6.3 Using ChatGPT and Claude Together
ChatGPT Claude
Initial idea brainstorm critique and
and simplify structure
Human researcher
verify, rewrite,
decide
Figure 3: A safe two-tool workflow for improving thinking without losing human control.
Do not use ChatGPT or Claude for
Do not use them to fabricate references, write assignments for students without learning, make
final ethical decisions, diagnose human subjects, or submit unverified content under your name.
7 ResearchRabbit: Seeing the Literature Network
ResearchRabbit is useful when participants already have one or more important papers. These papers
are called seed papers. From the seed papers, the tool helps discover related papers, earlier foundational
works, newer papers and author networks. Its strength is visual discovery of literature relationships.2
7.1 Key Terms
Seed paper: A starting paper that is directly related to your topic.
Seminal paper: An older influential paper that shaped the field.
Cutting-edge paper: A recent paper that shows new directions.
Citation network: A map showing how papers are connected through references and citations.
Research gap: A missing or under-explored area in the literature.
2
ResearchRabbit describes itself as a tool to find related papers, build citation maps and track research trends.
6
AI in Higher Education Research Dr. Shujaat Ali
Earlier papers Newer papers
background who cited it
Seed paper
Author network Possible re- Topic clusters
search gaps
Figure 4: ResearchRabbit helps participants move from one paper to a wider research map.
7.2 Activity: Build a Literature Map
20-minute practical activity
1. Select one relevant paper from your field.
2. Add it as a seed paper in ResearchRabbit.
3. Identify five related papers, two older papers and two recent papers.
4. Write one possible research gap in your own words.
5. Share with a partner: Which paper is most important and why?
8 Semantic Scholar: Searching Scholarly Literature
Semantic Scholar is a free AI-powered research tool for scientific literature. It helps scholars discover
relevant research by using AI to understand the meaning of scientific literature, not only keywords.3
8.1 How to Search Better
• Start with broad keywords, then narrow them.
• Use synonyms: for example, “AI in libraries”, “machine learning in academic libraries”, and “generative
AI in higher education”.
• Sort by relevance, year and citations.
• Check abstracts first, then download full text when available.
• Save papers into a reference manager or organized folder.
Prompt after finding papers
I found these five paper titles and abstracts. Please group them into themes, identify what each
paper studies, and suggest what is still missing. Do not add papers that are not in my list.
9 NotebookLM: Turning Documents into a Knowledge Base
NotebookLM is useful when participants have their own documents, such as PDFs, lecture notes, reports,
policies, articles or meeting minutes. Users can upload sources and then ask questions based on those
3
Semantic Scholar describes itself as a free AI-powered research tool for scientific literature, based at Ai2.
7
AI in Higher Education Research Dr. Shujaat Ali
sources. This makes it useful for librarians, teachers and researchers who need source-grounded answers.
Google presents NotebookLM as an AI research tool and thinking partner.4
9.1 Three-Part NotebookLM Model
Source documents Outputs
NotebookLM
PDFs, notes, summary,
source-grounded AI
reports Q&A, guide
Quality depends on source qual-
ity. If the uploaded documents
are weak, incomplete or outdated,
the answer will also be limited.
Figure 5: NotebookLM is best understood as a source-based research notebook.
9.2 NotebookLM Workflow for Teachers and Librarians
1. Upload source documents.
2. Ask for a plain-language summary.
3. Ask for key concepts and definitions.
4. Ask for questions for students or participants.
5. Ask for a handout, checklist or FAQ.
6. Check every important claim against the source.
Prompt for source-based study guide
Using only the uploaded sources, create a study guide for workshop participants. Include key
concepts, simple definitions, examples, five discussion questions, and a final checklist. Mark any
point that is not clearly supported by the sources.
10 Consensus: Evidence-Based Answers
Consensus is useful when a participant wants to know what research papers say about a specific question.
It searches and analyzes peer-reviewed literature and presents evidence-based summaries. Its role is not
to replace reading; it helps the participant reach relevant literature faster.5
10.1 Good Questions for Consensus
• Does formative feedback improve student learning outcomes?
• Is blended learning effective in higher education?
• What are the benefits of AI literacy training for faculty?
• Does library instruction improve students’ information literacy?
10.2 Consensus with ChatGPT and Claude
Use Consensus to locate evidence. Then use ChatGPT or Claude to explain, compare and simplify the
findings. The human researcher must still read key papers and check methodology.
4
Google NotebookLM official website, AI Research Tool and Thinking Partner.
5
Consensus describes itself as a tool to search and analyze peer-reviewed literature.
8
AI in Higher Education Research Dr. Shujaat Ali
Key papers
Consensus
Research question read abstracts
evidence search
and methods
ChatGPT
Human judgment / Claude
verify and write summarize
and compare
Figure 6: Consensus should be used as an evidence entry point, not as the final authority.
11 STORM: Topic Exploration and Knowledge Curation
STORM is a Stanford research prototype that generates Wikipedia-like reports on a topic and supports
interactive knowledge curation. Its value is early exploration, multi-perspective outline generation and
topic familiarization.6
11.1 When STORM is Useful
• At the beginning of a new topic.
• When you need a broad map before reading deeply.
• When preparing a seminar outline.
• When identifying subtopics and possible angles.
STORM caution
A STORM report is not a literature review by itself. Use it as an orientation tool. Verify claims
through scholarly databases, primary papers and official sources.
12 Jenni AI and Paperpal: Academic Writing Support
Jenni AI and Paperpal are academic writing tools. They can support drafting, paraphrasing, clarity,
grammar, structure and citation-related workflows. Paperpal describes itself as an academic writing
assistant for stages from first draft to submission, and Jenni describes itself as an academic writing and
research assistant.7
12.1 Where They Fit in the Writing Process
First draft
Research notes Outline
Jenni AI
Language edit
Final manuscript Human revision
Paperpal
Figure 7: Writing tools help improve drafts, but the researcher controls meaning and originality.
6
Stanford STORM official site and Stanford OVAL GitHub describe STORM as a research prototype for knowledge
curation and Wikipedia-like report generation.
7
Paperpal official website; Jenni AI official website.
9
AI in Higher Education Research Dr. Shujaat Ali
Prompt for improving a paragraph
Improve the following academic paragraph for clarity, flow and grammar. Keep the meaning
unchanged. Do not add new claims. After revision, list what you changed and why.
13 AI for Librarians
Librarians play a central role in responsible AI adoption because they support information literacy,
literature discovery, citation ethics and source evaluation. AI does not reduce the role of libraries; it
expands the need for trained information professionals.
13.1 AI-Supported Library Services
• Literature search clinics using Semantic Scholar and ResearchRabbit.
• Source evaluation sessions: credible, current, relevant and peer-reviewed.
• Prompt literacy training for students and faculty.
• Support for reference managers and citation checking.
• Institutional guides on responsible AI use.
• Workshops on avoiding fabricated references and plagiarism.
Search skills Citation ethics
Research capacity
University Library for faculty
and students
Prompt literacy Source
evaluation
Figure 8: The modern academic library can become a hub for responsible AI research support.
14 AI for Teachers and Supervisors
Teachers can use AI to design learning material, create examples, generate quizzes, explain difficult
concepts and give formative feedback. Supervisors can use AI to help students refine research questions,
structure proposals and improve writing. However, assessment must be designed so that students
demonstrate understanding, not only AI output.
14.1 Teaching Prompts
Prompt for lesson planning
Create a 60-minute lesson plan on [topic] for undergraduate students from mixed backgrounds.
Include learning outcomes, simple explanation, activity, assessment question and one real-life
example from Pakistan’s higher education context.
10
AI in Higher Education Research Dr. Shujaat Ali
Prompt for supervisor feedback
Act as a research supervisor. Review the following research proposal summary. Identify strengths,
weaknesses, unclear terms, missing literature, possible ethical issues and three next steps. Use
supportive language.
15 AI for Administrators and Academic Managers
AI can help administrators prepare reports, summarize meeting minutes, draft policy notes, analyze
feedback and plan training. But confidential data must be handled carefully. Sensitive student, staff or
institutional information should not be uploaded to public tools unless institutional policy allows it.
15.1 Administrative Uses
• Summarizing policy documents.
• Drafting minutes, circulars and training announcements.
• Analyzing anonymous feedback themes.
• Preparing checklists for accreditation or quality assurance.
• Designing staff development plans.
16 Responsible AI Use: VERIFY Model
The VERIFY model is a simple checklist for responsible academic use of AI.
V - Validate facts, numbers and references
E - Examine bias, missing context and overclaiming
R - Rewrite in your own academic voice
I - Identify what came from AI and what came from sources
F - Follow university, journal and ethics policies
Y - You are responsible for final content
Figure 9: VERIFY model for safe AI-supported research.
17 Academic Integrity and Ethics
17.1 Core Principles
1. Transparency: Disclose AI use when required by the journal, supervisor or institution.
2. Accountability: The author remains responsible for all claims.
3. Originality: Do not present AI-generated text as original thinking without revision and understanding.
4. Privacy: Do not upload confidential data without permission.
5. Fairness: Do not use AI in ways that disadvantage students or violate assessment rules.
11
AI in Higher Education Research Dr. Shujaat Ali
6. Human oversight: Use AI output as a draft or suggestion, not as final truth.
Confidentiality warning
Avoid uploading student records, unpublished data, examination papers, identifiable human-subject
data, confidential institutional files or private correspondence to public AI systems unless approved
by institutional policy and ethics requirements.
18 Prompt Library for Participants
18.1 Topic Selection
Prompt
I want to conduct research on [broad area]. Suggest five focused research topics suitable for
higher education research. For each topic, provide problem statement, possible objective, target
population, data collection method and expected contribution.
18.2 Literature Review Matrix
Prompt
Create a literature review matrix from the following paper details. Columns should include
author/year, objective, method, sample/data, key findings, limitation and relevance to my study.
Do not add information not provided.
18.3 Research Gap
Prompt
Based only on the summaries below, identify possible research gaps. Separate them into population
gap, methodology gap, context gap, theory gap and practical gap. Explain each in simple language.
18.4 Methodology Explanation
Prompt
Explain which methodology may fit my research question: [question]. Compare qualitative,
quantitative and mixed-method designs in a simple table. Include strengths, weaknesses and ethical
considerations.
18.5 Manuscript Editing
Prompt
Edit the following section for clarity, academic tone and logical flow. Do not change the meaning,
do not add unsupported claims, and list all changes made.
19 Hands-On Workshop Activities
Activity Tool Task Output
12
AI in Higher Education Research Dr. Shujaat Ali
1 ChatGPT / Claude Convert a broad topic into research 3 research questions
questions
2 Semantic Scholar Search papers and select relevant 5 paper list
literature
3 ResearchRabbit Build a citation map from one seed Visual paper cluster
paper
4 NotebookLM Upload sources and create a study Source-based notes
guide
5 Consensus Ask one evidence question Evidence summary
6 Jenni / Paperpal Improve a paragraph Revised paragraph
7 Group work Apply VERIFY model Ethical checklist
20 Assessment: Quick Reflection Questions
1. Which AI tool is most useful for your current work and why?
2. What is one AI risk you must manage in your role?
3. How can your department or library teach responsible AI use?
4. Which task should remain fully under human control?
5. How will you verify AI-generated output before using it?
21 Model Institutional Policy Points
A university or department can adopt the following simple AI policy points:
• AI may be used for brainstorming, planning, explanation and editing support.
• AI use must be acknowledged where required.
• AI must not be used to fabricate data, citations or results.
• Students must follow assessment-specific rules.
• Confidential data must not be uploaded to public tools.
• Final responsibility remains with the author, teacher, researcher or staff member.
22 Conclusion
AI tools can improve higher education research when used responsibly. The best approach is not to
ask, “Which AI tool can do my work?” Instead, ask, “Which AI tool can support my thinking while I
remain responsible for quality and ethics?” Teachers, librarians, researchers and administrators can all
benefit from AI, but only when they combine tool use with human judgment, disciplinary knowledge and
academic integrity.
13
AI in Higher Education Research Dr. Shujaat Ali
Selected Sources for Tool Descriptions
1. OpenAI. Introducing ChatGPT Edu. 2024. [Link]
du/
2. Anthropic. Introducing Claude for Education. 2025. [Link]
cing-claude-for-education
3. ResearchRabbit. AI Tool for Smarter, Faster Literature Reviews. [Link]
4. Semantic Scholar. AI-Powered Research Tool. [Link]
5. Google. NotebookLM: AI Research Tool and Thinking Partner. [Link]
6. Consensus. AI for Research. [Link]
7. Stanford STORM. [Link]
8. Jenni AI. AI Academic Writer and Research Tool. [Link]
9. Paperpal. AI Academic Writing Tool. [Link]
14