A practical guide for educators ready to transform their classrooms
December 2025 | Prof. Hiran Amarasekera | AI in Education
The AI Reality Every Academic Must Face
Walk into any university campus today, and you’ll see students chatting with ChatGPT between lectures. Visit any faculty meeting, and you’ll hear colleagues debating whether to ban these tools or embrace them.
Here’s the uncomfortable truth: 87% of students worldwide already use AI for assignments. Yet most universities are still figuring out what to do about it.
This guide shares five practical insights from faculty training sessions that will help academics move from confusion to confident action.
1. The “Catch the Cheater” Game is Already Over
Many academics spend hours trying to detect AI-generated submissions. Students spend minutes finding workarounds. This cycle helps no one.
The numbers tell the story:
- 87% of students use ChatGPT for academic work
- 62% use AI tools daily
- Only 23% ask for permission first
What happens if we keep ignoring AI:
- Students use these tools poorly, without guidance
- Academic integrity continues to decline
- Graduates enter workplaces unprepared for AI-integrated roles
What happens if we embrace AI:
- Students learn to use these tools ethically and effectively
- Critical thinking shifts to evaluating AI outputs, not just creating content
- Graduates become future-ready professionals
The shift: Stop being a “police officer” catching cheaters. Become a “guide” teaching students to use AI responsibly.
2. Ignoring AI is No Longer an Option
Every generation of academics has faced technological disruption:
YearTechnologyImpact1975MicrosoftPersonal computing1998GoogleInformation access2004FacebookSocial interaction2007iPhoneMobile computing2022ChatGPTContent generation
ChatGPT reached one million users in just five days — the fastest adoption of any technology in history.
Why academics must act now:
- It’s inevitable — Your students are already using AI. Their future employers expect AI proficiency.
- It’s beneficial — AI can handle routine tasks like marking schemes, lesson planning, and literature reviews.
- It’s urgent — Early adopters gain advantages. Late adopters struggle to catch up.
The question is no longer whether to integrate AI into teaching. The question is how quickly you can adapt.
3. Your Traditional Assignments Are Now Obsolete
The old model looked like this:
PowerPoint Lecture → Memorize Content → Final Exam → Forget Everything
This approach fails when AI can generate answers instantly. Assignments that test recall are now pointless.
The new model centres on real-world problems:
Real Problem → AI-Assisted Research → Critical Analysis → Practical Output
The Assignment Redesign Prompt
Use this prompt to transform any outdated assignment:
“Act as an academic instructional designer. I currently ask students to write a traditional assignment on: ‘[Insert Your Topic Here]’. Convert this into a problem-based, AI-assisted assignment that:
1. Encourages the use of tools like ChatGPT or Perplexity for research 2. Involves a real-world role (e.g., policy advisor, engineer, consultant) 3. Includes a scenario and a decision-making task 4. Results in a practical outcome (e.g., a report, action plan, memo)
Format your answer with a new assignment description, the role/scenario, expected output, and suggested AI prompts for students.”
Before and After Example
Old assignment (easily plagiarised):
“Write a 1000-word essay on ocean pollution.”
New assignment (AI-assisted, skill-building):
“You are an environmental advisor to your national fisheries ministry. Using AI tools for research, draft a strategic policy memo outlining three interventions to reduce marine plastic waste in your coastal region. Include cost estimates and implementation timelines.”
The student still researches ocean pollution — but now they develop professional skills, critical thinking, and practical document writing.
4. Prompt Engineering: The Skill Every Academic Needs
Why do AI tools sometimes give useless answers? Because vague questions produce vague responses.
AI is not Google. It doesn’t retrieve existing documents. It generates new content based on your instructions. Short prompts fail. Detailed prompts succeed.
The R-A-C-F-Q Framework for Better Prompts
LetterElementPurposeRRoleDefine the AI's perspectiveAActionSpecify the exact taskCContextProvide background informationFFormatDetail the desired structureQQuestionsAsk AI to clarify before starting
Example: A Weak Prompt vs A Strong Prompt
Weak prompt:
“Explain photosynthesis.”
Strong prompt using R-A-C-F-Q:
“Act as a senior lecturer in plant biology. I am preparing lesson notes for first-year undergraduates on plant physiology. Explain photosynthesis, highlighting the light and dark reactions, and common student misconceptions. Present the output as clear bullet points with real-world examples from agriculture. Ask me clarifying questions before you begin.”
The difference in output quality is dramatic. Teaching this framework to students transforms how they interact with AI — from passive consumers to active collaborators.
5. AI Won’t Do Your Research — But It Will Save You Hours
Every researcher knows the daily grind:
- Searching through millions of publications
- Reading dense PDFs to find one useful paragraph
- Organising notes scattered across folders
- Writing and rewriting academic prose
The insight: The struggle isn’t the thinking — it’s the administrative burden. AI can handle that burden.
The 5-Stage AI Research Workflow
| Stage | What AI Does | Recommended Tools |
|---|---|---|
| Discovery | Finds relevant papers using plain English questions | Elicit, Research Rabbit, Connected Papers |
| Understanding | Summarises PDFs, extracts key findings | ChatPDF, NotebookLM, Humata |
| Organisation | Categorises notes, identifies themes | Glasp, Obsidian, Lateral |
| Writing | Improves academic prose, suggests citations | Jenni, Paperpal, Writefull |
| Review | Provides feedback before peer review | Stanford Agentic Reviewer |
Important reminder: These tools augment your research. They don’t replace your critical analysis, interpretation, or original contribution. AI handles the mechanical work — you provide the insight.
What Will You Do Tomorrow?
The choice facing academics worldwide is clear:
Option A: Continue as before. Hope AI goes away. Watch students use these tools poorly. Graduate students unprepared for modern workplaces.
Option B: Embrace the change. Redesign assignments. Teach prompt engineering. Use AI to reduce administrative burden. Graduate students who can think critically AND use technology effectively.
Universities that adapt will produce graduates ready for tomorrow’s problems.
Universities that resist will produce graduates memorising yesterday’s answers.
Which future will you help build?
Quick Action Steps for This Week
- Try one AI tool — Start with Elicit for literature discovery or NotebookLM for understanding papers
- Redesign one assignment — Use the meta-prompt above to transform a traditional essay into a problem-based task
- Learn prompt engineering — Practice the R-A-C-F-Q framework with your own teaching materials
- Talk to your students — Ask them which AI tools they’re already using (the answers may surprise you)
Related Resources
- How AI Is Transforming Research: Essential Tools and How to Use Them
- Free AI tools for students and academics
- Prompt engineering templates for different disciplines
Prof. Hiran Amarasekera is a Professor at the University of Sri Jayewardenepura and an AI adoption consultant who conducts workshops on AI integration for universities and businesses.
Website: StudentLanka.com | TikTok: @hiranamara | Facebook @hiranbiology
Leave a Reply