For information and reference only – not to be submitted as your own work.
This sample shows how a Masters-level information studies student might approach a values, ethics and IPR written report that combines a ChatGPT query, a 1,000-word commentary and a 2,000–3,000-word essay. It is based on the brief behind this live question page, and the extract below is a ChatGPT essay commentary sample built around one question: should works generated by AI be protected by copyright? If you want the marking logic behind it, read our guide on how to critique a ChatGPT essay and go beyond it. For general essay structure support, see our essay writing service page.
What this sample covers
The brief has four linked deliverables, and the sample follows the same order. Because the full report would run to roughly 4,500 words, this page gives an extract of each part rather than a complete submission.
- The chosen question, why it was chosen and a short overview of the answer.
- The query written to ChatGPT, with the reasoning behind its wording.
- An extract of the commentary on the AI-generated essay (accuracy, references, gaps, other observations).
- The outline and opening section of the student’s own essay, showing how it extends the AI output.
The brief states no mark split, so none is claimed here. Check your own course handbook for weightings.
Part 1: Chosen question and why
Question: Should works generated by AI be protected by copyright, and if so, who should be treated as the author?
Why this question. It sits exactly where the Values & IPR lecture material meets current practice. Copyright is built around human creativity, yet generative systems now produce text and images that look like authored work. The question is also answerable: courts and regulators have started to take positions, so the essay can argue from evidence rather than speculation. Finally, it has an ethical side (who benefits, who is credited, whose work trained the system) as well as a legal side, which suits an information studies audience.
Overview of the answer. The essay argues that, on current law in the jurisdictions examined, output with no meaningful human creative contribution should not attract copyright, but that human selection, arrangement and editing of AI output can. It then argues that the harder policy problem sits on the input side: how training data is copied and used. The essay treats these as two separate questions, which the ChatGPT draft did not.
Part 2: The query written for ChatGPT
A weak query asks for “an essay on AI and copyright”. A better one tells the tool what you will later be assessed on, so the output is a useful test object. The example below is one a student could use.
“Write a 1,200-word essay answering the question: should works generated by artificial intelligence be protected by copyright, and who should be regarded as the author? Cover the concept of authorship, the position under the law of at least two jurisdictions, and the ethical arguments for and against protection. Use an academic tone and include in-text citations and a reference list in APA 7 style.”
Three choices are deliberate. The word count keeps the output short enough to annotate. Asking for two jurisdictions forces claims that can be checked. Asking for APA citations invites exactly the references you will need to test in the commentary. Record the date and the tool version used, and keep the full output as an appendix.
Part 3: Commentary extract (about 700 words)
Note: the AI text below is not a real transcript. It is an illustrative paraphrase of the kind of output a general-purpose chatbot typically produces for this prompt.
Overview of the AI output
The generated essay opens with a general statement that AI is transforming creativity, defines authorship in one sentence, and then moves through “arguments for” and “arguments against” protection in two balanced paragraphs. It concludes that “a balanced approach is needed”. It cites four or five sources in APA format. This structure is typical: tidy and confident, but descriptive rather than argued.
Accuracy
The broad claim that most legal systems tie copyright to human authorship is directionally right for the United States. In Thaler v. Perlmutter, the D.C. Circuit held on 18 March 2025 that human authorship is required for registration and that an AI system cannot be an author (Thaler v. Perlmutter, 2025). The US Copyright Office’s January 2025 report goes further and says that, with current technology, prompts alone do not give users enough control to count as authors, although human expressive contributions and creative arrangement of output can be protected (U.S. Copyright Office, 2025).
The weakness in the AI draft is precision. A typical output says “courts have ruled that AI cannot hold copyright” without naming the case, the court, the year or the narrow point decided. That sentence is also too broad: the case concerned registration where the machine was listed as sole author, not every use of AI in creating a work. A commentary should say so. A second accuracy problem is jurisdictional drift. AI essays often present US doctrine as if it were general law, and say little or nothing about Singapore, where the Copyright Act 2021 governs.
Nature of the references
The reference list is the part of the AI essay that most needs checking line by line. Research on ChatGPT-generated citations found that many fabricated references, with plausible authors, journal names and article titles that could not be matched to a real publication, and many real ones contained errors (Walters & Wilder, 2023). So each reference should be searched in a library database or through Crossref, and the commentary should report what was found: how many exist, how many have the details right, and whether any supports the claim it is attached to. In the illustrative case here, the reference list would be judged on three tests: existence, accuracy of details, and relevance. A reference that exists but concerns a different topic still fails the third test.
The type of source matters too. The AI draft leans on general commentary and cites no primary law, no court decision and no regulator report. A Masters essay should rest on statute, case law and peer-reviewed work, with commentary used to interpret them.
Gaps and omissions
- Inputs versus outputs. The AI essay blurs ownership of generated output with the legality of copying works to train a model. These are different questions with different answers.
- Singapore. The computational data analysis exception in the Copyright Act 2021 allows copying for analysis under conditions, including lawful access to the source material (Tan, 2021). The AI draft does not mention it.
- The spectrum of human involvement. The draft treats “AI-generated” as one category. Between a single prompt and heavy human editing sits a range, and the answer changes along it.
- Ethical depth. Values are listed (fairness, creativity, innovation) but not applied to anyone: creators whose work was used for training, users, or publishers.
Other observations
The essay is fluent and uniformly confident, even where it is wrong. It states no limits to its own knowledge and does not flag that law in this area is still changing. It also has no real thesis, which is the main reason it would score modestly as an essay in its own right. That is the useful lesson for the commentary: the quality of prose is not evidence of the quality of analysis.
Hire a Professional Essay & Assignment Writer for completing your Academic Assessments
Native Singapore Writers Team
- 100% Plagiarism-Free Essay
- Highest Satisfaction Rate
- Free Revision
- On-Time Delivery
Part 4: Own essay, outline and opening section
Outline (about 2,500 words)
| Section | Purpose | Approx. words |
|---|---|---|
| Introduction and thesis | Define the question; argue that protection should follow human creative contribution | 300 |
| Authorship and values | Why copyright tied authorship to people; incentive and personality rationales | 450 |
| Outputs: law in practice | US decision and Copyright Office guidance; the position of prompts | 600 |
| Inputs: copying for training | Singapore’s computational data analysis exception and its limits | 500 |
| Ethical analysis | Creators, users, and the public domain | 450 |
| Conclusion | Answer the question and name remaining uncertainty | 200 |
Opening section (extract)
Copyright law was designed around people. The reason is not only doctrinal. Protection exists to reward and encourage human creative effort, and that rationale weakens when no person has exercised creative judgement over the result. This essay argues that copyright in AI-generated material should depend on the degree of human creative contribution, not on the tool used, and that the more serious policy conflict lies in how works are copied to train such systems.
The argument starts from the current legal position on outputs. In the United States, the D.C. Circuit has affirmed that a work with a machine as its only listed author cannot be registered, because the Copyright Act requires human authorship (Thaler v. Perlmutter, 2025). That decision is narrow. It does not say that works made with AI assistance are unprotectable. The Copyright Office’s analysis draws the line more finely: prompts alone are insufficient, but a human’s original expression that is visible in the output, or a creative selection and arrangement of generated material, may be protected (U.S. Copyright Office, 2025).
Where the ChatGPT draft stopped at “a balanced approach is needed”, this essay takes a position and tests it against inputs as well as outputs. On inputs, Singapore’s Copyright Act 2021 permits copying a work for computational data analysis under stated conditions, including lawful access to the source (Copyright Act 2021; Tan, 2021). Whether that exception sufficiently respects the interests of creators is a values question, and it is developed in the ethical analysis section.
One discipline applies throughout: every source in this essay was located and read in full by the student, which is the standard the commentary applies to the AI output. Where this essay uses an idea that originated in the ChatGPT draft, such as the pairing of incentive and personality rationales, it is restated in the student’s own words and then developed with scholarship the AI draft did not cite. The ChatGPT output is a starting point, never a source of text.
Further help
If you are planning the research stage, our research paper help page explains how we support literature searching and source evaluation, and the information governance assignment help page covers related policy and compliance topics. Use any sample as a model for structure and reasoning only, and follow your institution’s rules on AI use and academic integrity.
About the expert
According to his profile, Mark Wong holds a Master’s degree in Journalism and Mass Communication from the University of Wisconsin-Madison, and his areas include media law and ethics, public relations, journalism and advertising.
References
Copyright Act 2021 (Singapore). https://sso.agc.gov.sg/Act/CA2021
Tan, D. (2021, December 15). “We don’t need permission to dance”: Key features of the new Copyright Act 2021. The Singapore Law Gazette. https://lawgazette.com.sg/feature/we-dont-need-permission-to-dance/
Thaler v. Perlmutter, No. 23-5233 (D.C. Cir. Mar. 18, 2025).
U.S. Copyright Office. (2025). Copyright and artificial intelligence, part 2: Copyrightability. https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf
Walters, W. H., & Wilder, E. I. (2023). Fabrication and errors in the bibliographic citations generated by ChatGPT. Scientific Reports, 13, Article 14045. https://doi.org/10.1038/s41598-023-41032-5
