ES1103 Chap.10 Researching Responsibly with Generative AI
Researching Responsibly with Generative AI
AI inside the course, not outside it
Generative AI appears in this course as a subject to be learned rather than a rule to be obeyed. The first tutorial discusses the advantages and disadvantages of using it for academic work and what ethical, responsible use means.
The fifth tutorial has students brainstorm contentious issues themselves, then put the same prompt to several different AI tools and compare the outputs with the human list.
The podcast for the first essay states the course's position simply: AI should be a support and not a substitute, and students should lead every step of their own writing process.
Principles for research with AI
The fifth tutorial sets out principles with concrete practices.
Keep a detailed record of use, including the tools, the exact prompts or a link to the chat log, and how the output was changed and integrated. Prioritise verification against original sources to guard against fabricated or distorted information, checking credibility and representation. Use AI to scaffold inquiry rather than replace it, for example by asking it to critique your work or explain a concept with evidence.
And regulate your own process by asking whether you agree with the output, whether it needs validating in a research database, whether it should change your focus, and what you will do if it is wrong.
Four kinds of question
To engage critically with AI output the tutorial adapts four question types from Pohlman and Thomas (2015). Clarifying questions improve understanding of what was generated.
Funnelling questions probe assumptions, evidence and causes. Adjoining questions explore related perspectives and stakeholders. Elevating questions step back to implications, limitations and future developments.
Used in sequence, they turn a chatbot's first answer into the start of an inquiry rather than its end.
Declaring use
Each essay task requires the course declaration to be pasted at the end of every submission. It states whether generative AI was used and, where it was, records the tool, the prompt and output, and how the output was used.
The university's policy on AI in learning explains that acknowledged use is not plagiarism and is not penalised for being declared, while unacknowledged use is treated like copying or paraphrasing another author without credit. It also leaves responsibility for the accuracy and integrity of the submission with the student. Some in-class activities, such as the timed paragraph in the second tutorial, do not allow AI at all.
What this chapter covers
- 01
How the course positions AI: support, not substitute
- 02
Recording prompts, outputs and changes as you work
- 03
Verifying AI claims against original sources
- 04
Clarifying, funnelling, adjoining and elevating questions
- 05
The declaration and the university policy behind it
Worked example · free
Auditing a chatbot's research suggestion
- 2State the verification step.
- 2Give one funnelling and one elevating question.
- 1State what goes in the declaration.
Key terms
- AI Declaration
- The statement at the end of each submission recording whether AI was used and, if so, the tool, prompt, output and how it was used.
- AI Hallucination
- Output from a generative AI tool that presents fabricated or inaccurate information as fact.
- Funnelling Question
- A question that probes the assumptions, evidence and causes behind an AI output.
- Elevating Question
- A question that steps back to ask about implications, limitations and the wider significance of an issue.
- Self-Regulation
- Monitoring and adjusting your own thinking process, for example by questioning whether you agree with AI output.
- Source Verification
- Checking a claim against the original source for accuracy, credibility and faithful representation.
- Adjoining Question
- A question that explores related perspectives, stakeholders or contexts around an issue.
Researching Responsibly with Generative AI FAQ
Is using generative AI allowed in the essays?
The course teaches responsible use rather than banning it, and requires a declaration at the end of every submission. The university policy states that acknowledged use is not plagiarism and is not penalised for being declared, but you remain responsible for accuracy and integrity.
What must the AI declaration contain?
A statement of whether you used generative AI and, for each use, the tool, your prompt and its output, and how the output was used in the assignment. The fifth tutorial recommends recording exact prompts or a link to the chat log as you work.
Why should I not rely on AI to summarise sources?
Tutors report that AI summaries of the course readings have repeatedly contained inaccuracies, and the course's principles ask you to verify against original sources. A summary you have not checked can misrepresent a source, which the rubric penalises.
What is the difference between clarifying and elevating questions?
Clarifying questions make sure you understand what the AI said, for example by asking for simpler wording or an example. Elevating questions move beyond it to implications, limitations and future developments, which is where material for an argument often comes from.
Where should I verify what a chatbot tells me?
In the original source, found through the university library's search. The fifth tutorial's homework practises exactly this: pick one point from an AI conversation, search for it, and check whether the source says what the chatbot claimed.
Assessment move
Open a running log the first time you use AI for this course and add a line every time: date, tool, prompt, what you kept and what you checked. When a submission is due, the declaration is already written, and the log becomes evidence of your process for the final conversation.
Working through Researching Responsibly with Generative AI in ES1103? Sia is AskSia’s AI Arts and Humanities tutor — ask any ES1103 Researching Responsibly with Generative AI question and get a clear, step-by-step explanation grounded in how ES1103 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
