The DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index; none of the 18 sources its claims cite is checked for support yet.
ChatGPT and large language models in academia: opportunities and challenges
Open accessAbstract
The introduction of large language models (LLMs) that allow iterative “chat” in late 2022 is a paradigm shift that enables generation of text often indistinguishable from that written by humans. LLM-based chatbots have immense potential to improve academic work efficiency, but the ethical implications of their fair use and inherent bias must be considered. In this editorial, we discuss this technology from the academic’s perspective with regard to its limitations and utility for academic writing, education, and programming. We end with our stance with regard to using LLMs and chatbots in academia, which is summarized as (1) we must find ways to effectively use them, (2) their use does not constitute plagiarism (although they may produce plagiarized text), (3) we must quantify their bias, (4) users must be cautious of their poor accuracy, and (5) the future is bright for their application to research and as an academic tool.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
This editorial argues that LLM-based chatbots such as ChatGPT should be embraced in academia for writing, education, and programming, while cautioning about bias, hallucination, and poor accuracy. The authors conclude that LLM use does not constitute plagiarism, that bias metrics are needed, and that users remain responsible for validating outputs.
Key findings
The authors’ stance is that (1) LLMs must be embraced, (2) their use does not constitute plagiarism although they may produce plagiarized text, (3) their bias must be quantified, (4) users must be cautious of poor accuracy, and (5) the future is bright for their application to research and as an academic tool.
- ChatGPT can present false information as truth: when asked to help find relevant publications to cite, it “proceeded to make up a list of five entirely fictional publications”; on repeat trials it sometimes listed one or two real papers, but other times fabricated a paper or returned an irrelevant one.
- On scalability and accessibility, the paper reports ChatGPT gained 100 million active users in 2 months, with an estimated 13 million daily queries, compared with Google search’s estimated 8.5 billion daily queries; ChatGPT offers a free tier and a $20/month premium plan. It also notes BLOOM as an open access 176B parameter multilingual LLM, ChatGPT as a 175 billion parameter LLM, speculation that OpenAI had access to 10,000 Nvidia V100s costing close to $5 million, and an estimate that training BERT (described as a 6 billion parameter LLM) would produce 21 to 78 metric tons of CO2.
- For academic writing, LLMs may particularly help non-native English speakers; the paper cites that 49% of postdoctoral fellows trained in the United States were born overseas, and in engineering, math, and computer science, 60% of PhDs were awarded to international students. For programming, ChatGPT can help write, interpret, debug, simplify, and translate code, but it may present incorrect code as correct, miss edge cases, and is more reliable for smaller blocks than larger complex programs.
Methods
Design: editorial/commentary, not an empirical study or systematic review.
- Data or sample: no primary data; the paper draws on published literature, illustrative examples, and the authors’ experiences.
- Main analysis: narrative discussion of opportunities and challenges across academic writing, education, programming, bias, and scalability/accessibility.
Limitations
The paper does not report a formal limitations section for its own approach. Evident from the text: it is an editorial/commentary, so its examples are illustrative and include personal experiences rather than systematic evidence.
- The paper states LLM limitations: hallucination/false information, bias from training data, poor accuracy, misrepresentation of capabilities, and inability to anticipate edge cases in code.
- The paper notes unresolved ethical/legal uncertainty around LLM use in academic writing and grants, and that detection of cheating remains an unsolved challenge for educators.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
17 Own interpretation · 9 Not checked yet · 7 No basis given · 4 Own result
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A fundamental challenge with LLM-based chatbots is that they can present false information as truth, also known as “hallucination”.”
- Also stated in: § Background · plain
“For example, when asked if it could help find relevant publications to cite in a review paper, ChatGPT confirmed that it could, but then proceeded to make up a list of five entirely fictional publications.”
§ ChatGPT and large language models in academia: opportunities and challenges > Background
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Background · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In fact, ChatGPT and Bard can misrepresent their capabilities.”
§ ChatGPT and large language models in academia: opportunities and challenges > Background
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“For example, LLMs may adopt bias, perpetuate stereotypes in the training dataset, and present false information as truth.”
§ ChatGPT and large language models in academia: opportunities and challenges > Introduction
What the sources say
- Not checked yetcites [3] Sallam M. ChatGPT utility in healthcare education, research, and practice: systematic review on the promising perspectives and valid concerns. Healthcare. 2023;11(6):887. https://doi.org/10.3390/healthcare11060887. source
- Not checked yetcites [4] Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepaño C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digit Health. 2023;2:e0000198. source
- Not checked yetcites [5] Liebrenz M, Schleifer R, Buadze A, Bhugra D, Smith A. Generating scholarly content with ChatGPT: ethical challenges for medical publishing. Lancet Digit Health. 2023;5:e105–6. source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These issues limit the utility of LLM chatbots in some domains, such as the medical field, where facts are essential to ensure the best health outcomes.”
§ ChatGPT and large language models in academia: opportunities and challenges > Background
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus, an answer generated by a large language model can be formatted correctly but not necessarily be factual.”
§ ChatGPT and large language models in academia: opportunities and challenges > Background
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“Over reliance may hinder students ability to develop critical skills, such as writing [21].”
§ ChatGPT and large language models in academia: opportunities and challenges > Education
What the sources say
- Not checked yetcites [21] Shidiq M. The use of artificial intelligence-based chat-gpt and its challenges for the world of education; from the viewpoint of the development of creative writing skills. Proc Int Conf Educ Soc Humanity. 2023;1(1):353–7. ISSN 2986-5832.
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Consequently, employing LMMs with pre-written text is unlikely to raise ethical concerns such as plagiarism, which may arise when using prompt-based text generation by LMMs.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT as a writing and editing tool
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“Unethically and irresponsibly designed AI/ML solutions deployed in the healthcare setting can exacerbate and perpetuate systematic biases and disparities for those from marginalized groups [27–29].”
§ ChatGPT and large language models in academia: opportunities and challenges > Bias
What the sources say
- Not checked yetcites [27] Vyas DA, Eisenstein LG, Jones DS. Hidden in Plain Sight - Reconsidering the Use of Race Correction in Clinical Algorithms. N Engl J Med. 2020;383:874–82. source
- Not checked yetcites [28] Gijsberts CM, Groenewegen KA, Hoefer IE, Eijkemans MJC, Asselbergs FW, Anderson TJ, et al. Race/Ethnic Differences in the Associations of the Framingham Risk Factors with Carotid IMT and Cardiovascular Events. PLoS ONE. 2015;10:e0132321. source
- Not checked yetcites [29] Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting Racial Bias in an Algorithm used to Manage the Health of Populations. Science. 2019;366:447–53. source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“A recent study estimated that the training of BERT (a 6 billion parameter LLM) would produce between 21 to 78 metric tons of CO2 to train to completion [14].”
§ ChatGPT and large language models in academia: opportunities and challenges > Scalability and accessibility of LLMs
What the sources say
- Not checked yetcites [14] Dodge J, Prewitt T, Des Combes RT, Odmark E, Schwartz R, Strubell E, Luccioni AS, Smith NA, DeCario N, Buchanan W. Measuring the Carbon Intensity of AI in Cloud Instances. arXiv. 2022;2206.05229. source
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“Until the root-causes of the encoded biases of LLMs are addressed, LLMs for clinical applications will suffer the same fate, propagating biases [30, 31].”
§ ChatGPT and large language models in academia: opportunities and challenges > Bias
What the sources say
- Not checked yetcites [30] Neeley T, Ruper S. Timnit Gebru: 'SILENCED No More' on AI Bias and The Harms of Large Language Models. Harvard Business School Case. 2022;422–085.
- Not checked yetcites [31] Bender EM, Gebru T, McMillan-Major A, Shmitchell S. On the dangers of stochastic parrots: can language models be too big? In: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency. New York, NY, USA: ACM; 2021. p. 610–23. source
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: unclear — It is unclear whether it could be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Probably not a substantial claim. 23
- It may state two separate claims as one. 11
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In one author's experience experimenting with how ChatGPT might help write grants, ChatGPT can help provide some ideas for interesting directions, even suggesting sound logical reasons that we need to do research that could find their way into the significance section of NIH grants.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT and LLMs for grant proposal writing
plaindirectSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, grants must convey many complex ideas in a short amount of space, and as such, since much of the text from ChatGPT is so general and often repetitive, in our experience none of the text produced by ChatGPT made it into the final grant proposal.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT and LLMs for grant proposal writing
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“ChatGPT can even provide general text that might be logically used in parts of the grant, such as the aims page.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT and LLMs for grant proposal writing
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“A more important concern regarding students learning from a LLM is that it typically responds in a confident, and seemingly knowing manner, whether the provided information is accurate or not.”
§ ChatGPT and large language models in academia: opportunities and challenges > Education
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It can suggest specific experiments that might test a hypothesis of interest.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT and LLMs for grant proposal writing
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“A surprising effect for one author of this editorial was that it helped overcome writer's block in that simply having some text on the page was sufficient to move forward with editing.”
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > ChatGPT and LLMs for grant proposal writing
plaindirectSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The model can generate correctly formatted information such as references, that do not exist.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“ChatGPT will misrepresent its knowledge; instead of responding “I don’t know”, it will readily provide fabricated information with apparent confidence.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“For example, LLMs may adopt bias, perpetuate stereotypes in the training dataset, and present false information as truth.”
§ ChatGPT and large language models in academia: opportunities and challenges > Introduction
What the sources say
- Not checked yetcites [3] Sallam M. ChatGPT utility in healthcare education, research, and practice: systematic review on the promising perspectives and valid concerns. Healthcare. 2023;11(6):887. https://doi.org/10.3390/healthcare11060887. source
- Not checked yetcites [4] Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepaño C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digit Health. 2023;2:e0000198. source
- Not checked yetcites [5] Liebrenz M, Schleifer R, Buadze A, Bhugra D, Smith A. Generating scholarly content with ChatGPT: ethical challenges for medical publishing. Lancet Digit Health. 2023;5:e105–6. source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, it would not be wise to base treatment plans on outputs from ChatGPT, at least in its current state, because this would lead to a large proportion of misdiagnoses and ineffective treatments that could cause patient harm.”
§ ChatGPT and large language models in academia: opportunities and challenges > Background
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Deploying AI models at scale is a challenge that future AI technologies will have to face even for tech giants such as Google (Bard power by LaMDA) and Microsoft (AI powered Bing) that have announced AI integration within their respective search engines.”
§ ChatGPT and large language models in academia: opportunities and challenges > Scalability and accessibility of LLMs
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In line with warnings for general use, ChatGPT may somewhat unpredictably present incorrect code as being correct.”
§ ChatGPT and large language models in academia: opportunities and challenges > Assistance with programming
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
In the paper
“Concerns have arisen regarding how students might easily use ChatGPT dishonestly, either to cheat in the completion of homework or exams, or to write reports/essays without citation in a manner that could be construed as plagiarism [22].”
§ ChatGPT and large language models in academia: opportunities and challenges > Education
What the sources say
- Not checked yetcites [22] Neumann M, Rauschenberger M, Schön E-M. “We Need To Talk About ChatGPT”: The Future of AI and Higher Education. 2023:4. https://doi.org/10.25968/opus-2467. source
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Currently, ChatGPT is more likely to be successful in accurately writing smaller blocks of code, whereas its reliability in writing larger/more complex programs (e.g. a software package) is questionable.”
§ ChatGPT and large language models in academia: opportunities and challenges > Assistance with programming
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Ultimately, code written by ChatGPT may offer a useful starting point, however both the code and comments that it generates should be checked and validated by both the programmer and users to confirm that it fully satisfies the intended purpose.”
§ ChatGPT and large language models in academia: opportunities and challenges > Assistance with programming
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“ChatGPT also promises to facilitate computer programming in terms of (1) learning to code or use specific packages, libraries, or frameworks, (2) writing new code, (3) interpreting existing code, (4) debugging existing code, (5) increasing the compactness or efficiency of code, or (6) translating code from one programming language to another.”
§ ChatGPT and large language models in academia: opportunities and challenges > Assistance with programming
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“LLMs provide benefits and opportunities in education for students by assisting with research and academic writing [15] or using them as an interactive study guide which may include generating practice exams and being provided immediate feedback [16, 17].”
§ ChatGPT and large language models in academia: opportunities and challenges > Education
What the sources say
- Not checked yetcites [15] Kasneci E, Sessler K, Küchemann S, Bannert M, Dementieva D, Fischer F, et al. ChatGPT for good? On opportunities and challenges of large language models for education. Learn Individ Differ. 2023;103:102274. source
- Not checked yetcites [16] MacNeil S, Tran A, Mogil D, Bernstein S, Ross E, Huang Z. Generating Diverse Code Explanations using the GPT-3 Large Language Model. In: Vahrenhold J, Fisler K, Hauswirth M, Franklin D, editors. Proceedings of the 2022 ACM Conference on International Computing Education Research - Volume 2. New York, NY, USA: ACM; 2022. p. 37–9. source
- Not checked yetcites [17] Tate TP, Doroudi S, Ritchie D, Xu Y, Mark Warschauer UCI. Educational Research and AI-Generated Writing: Confronting the Coming Tsunami. 2023.
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While AI has the potential to revolutionize education and academic research, paywalled access could further increase the divide between the wealthy and the poor.”
§ ChatGPT and large language models in academia: opportunities and challenges > Scalability and accessibility of LLMs
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
- Rests on an assumption: AI has the potential to revolutionize education and academic research.
In the paper
“Data produced from ChatGPT may be inaccurate or biased [23], which users need to be aware of to ensure these deficiencies are not propagated in their works.”
§ ChatGPT and large language models in academia: opportunities and challenges > Education
What the sources say
- Not checked yetcites [23] Baidoo-Anu D, Owusu Ansah L. Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning. 2023. Available at SSRN: https://ssrn.com/abstract=4337484 or https://doi.org/10.2139/ssrn.4337484. source
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This editorial provides several examples and perspectives on how LLMs increase the efficiency of academic writing, education, and programming.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In many contexts, LLMs have low accuracy.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Output from ChatGPT will match existing text on the internet, particularly if the sample is small enough.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
- Rests on an assumption: The output of ChatGPT is from most publicly available text on the internet until 2021.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, ChatGPT does not do anything without a prompt.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When LLMs for clinical applications are socially, consciously, and transparently designed with such considerations, they can become an additional tool for promoting equity and improving access to care.”
§ ChatGPT and large language models in academia: opportunities and challenges > Bias
plainconditional- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The performance of ChatGPT and other LLMs therefore mimics the available text, and they are as biased as their training data.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may state two separate claims as one.
- Rests on an assumption: The output of ChatGPT is from most publicly available text on the internet until 2021.
- Rests on an assumption: ChatGPT is known to perpetuate stereotypes such as nurses being female and doctors being male, and this bias comes from the training data.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus, in a strict sense, LLMs like ChatGPT do produce plagiarism.”
§ ChatGPT and large language models in academia: opportunities and challenges > Conclusion
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
- Rests on an assumption: The output of ChatGPT is from most publicly available text on the internet until 2021.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As companies race to innovate with LLMs like ChatGPT and others, encoded biases will be amplified, and harm will manifest.”
§ ChatGPT and large language models in academia: opportunities and challenges > Bias
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: fails — It may be too vague to check against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds. It may state two separate claims as one.
The paper relies on this without support.
Effective academic writing is essential to the success and quality of academic publications and is a major challenge facing international postdocs and students.
“Effective academic writing plays an essential role in the success and quality of academic publications, nevertheless, it is one of the major challenges facing international postdocs and students.”
The paper relies on this without support.
The current grant review system is subjective; some proposals scored very poorly in one study section while nearly the same proposal at another study section is funded.
“Despite this massive investment, the current review system is subjective; for example, some proposals sent to one study section are scored very poorly, while nearly the same proposal at another study section is funded.”
The paper relies on this without support.
ChatGPT has already been applied to help write grant proposals.
“ChatGPT has undoubtedly already been applied to help write grant proposals.”
The paper relies on this without support.
ChatGPT is a 175 billion parameter LLM.
“As a reminder, ChatGPT is a 175 billion parameter LLM and this only considers the cost of training and not any cost related to generating text in response to potentially billions of daily queries.”
The paper relies on this without support.
The output of ChatGPT is from most publicly available text on the internet until 2021.
“The output of ChatGPT is from most publicly available text on the internet until 2021.”
The paper relies on this without support.
ChatGPT is known to perpetuate stereotypes such as nurses being female and doctors being male, and this bias comes from the training data.
“For example, ChatGPT is known to perpetuate stereotypes such as nurses being female and doctors being male, and this bias comes from the training data.”
The paper relies on this without support.
The use of ChatGPT and other LLMs by students is inevitable.
“As the use of ChatGPT and other LLMs by students is inevitable, educators will need to change the pedagogy and assessments in education, such as integrating the use of ChatGPT in lessons [26], and a focus more on the process rather than the final product for assessment.”
The paper relies on this without support.
Detection of cheaters remains an unsolved challenge for educators.
“While the detection of cheaters remains an unsolved challenge for educators, if we start from the assumption that students are allowed, if not encouraged, to utilize ChatGPT, strategies could be adopted to account for this.”
The paper relies on this without support.
It is difficult to define metrics that assess the level of bias in training data and model outputs.
“It is difficult to define metrics that assess the level of bias in the training data and in the model outputs.”
The paper relies on this without support.
AI has the potential to revolutionize education and academic research.
“While AI has the potential to revolutionize education and academic research, paywalled access could further increase the divide between the wealthy and the poor.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“using GenAI to automate routine administrative tasks in higher education, such as consolidation of grades, creating performance summaries of students, crafting out evaluation reports, and maintaining student enrolment and attendance records (Meyer et al., 2023).”
From Higher Education's Generative Artificial Intelligence Paradox: The Meaning of Chatbot Mania (2024) · cites this paper as Meyer et al., 2023
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Third, LLMs can act as a thought partner for students and enhance understanding by providing detailed explanations and step-by-step solutions (Mollick and Mollick 2022, Kasneci et al. 2023, Meyer et al. 2023, Extance 2023).”
From AI Meets the Classroom: When Do Large Language Models Harm Learning? (2024) · cites this paper as Meyer et al. 2023
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“there has been a notable upswing in discussions among scholars and academicians regarding the transformative potential of generative AI, particularly ChatGPT, in higher education and the risks associated with it (Dalalah & Dalalah, 2023; Meyer et al., 2023; Peters et al., 2023; Yilmaz & Yilmaz, 2023a).”
From Is it harmful or helpful? Examining the causes and consequences of generative AI usage among university students (2024) · cites this paper as Meyer et al., 2023
In the citing paper
“The consequence of a self-referential loop in LLMs can lead to several problematic outcomes, including the propagation of biases [30], increased homogeneity in generated data, and ultimately, hindered performance.”
From Large Language Models and User Trust: Consequence of Self-Referential Learning Loop and the Deskilling of Health Care Professionals (2024) · cites this paper as [30]
In the citing paper
“ChatGPT is considered the fastest-growing technology in history: in just three months following its public launch, it amassed an estimated 120 million monthly active users [16] with an estimated 13 million daily queries [49], surpassing all other applications [64].”
From A systematic literature review of empirical research on ChatGPT in education (2024) · cites this paper as [49]
In the citing paper
“a fundamental challenge with LLM-based chatbots, including ChatGPT, is the accuracy and quality of the provided information and responses, as they provide false information as truth—a phenomenon often referred to as “hallucination” [3, 49]”
From A systematic literature review of empirical research on ChatGPT in education (2024) · cites this paper as [49]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ ChatGPT and large language models in academia: opportunities and challenges > ChatGPT for academic writing > Ethics and rules of using text from ChatGPT
“In one case, ChatGPT was listed as an author in the preprint [9] but missing as an author in the final publication [4].”
§ ChatGPT and large language models in academia: opportunities and challenges > References
“9. Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digital Health. 2023;2(2):e0000198. 10.1371/journal.pdig.0000198. PMC9931230 36812645”
Citations
Cites
ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · doi:10.35542/osf.io/5er8f · Cited by 7,552
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · cited 1×
“LLMs provide benefits and opportunities in education for students by assisting with research and academic writing [15] or using them as an interactive study guide which may include generating practice exams and being provided immediate feedback [16, 17].”
BackgroundDissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · cited 1×
“Unethically and irresponsibly designed AI/ML solutions deployed in the healthcare setting can exacerbate and perpetuate systematic biases and disparities for those from marginalized groups [27–29].”
BackgroundOn the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988
Kung, T. H., Cheatham, M., Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., Madriaga, M., Aggabao, R., Diaz-Candido, G., Maningo, J., et al. (2023) · PLOS Digital Health · doi:10.1371/journal.pdig.0000198 · Cited by 4,115 · cited 1×
“In one case, ChatGPT was listed as an author in the preprint [9] but missing as an author in the final publication [4].”
BackgroundBAİDOO-ANU, D., & OWUSU ANSAH, L. (2023) · Journal of AI · doi:10.61969/jai.1337500 · Cited by 3,939
Sallam, M. (2023) · Healthcare · doi:10.3390/healthcare11060887 · Cited by 3,126 · cited 1×
“On the other hand, there are also unsettling ethical issues to be considered [3–5].”
BackgroundChatting and cheating: Ensuring academic integrity in the era of ChatGPT
Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024) · Innovations in Education and Teaching International · doi:10.1080/14703297.2023.2190148 · Cited by 3,087
Chatting and Cheating. Ensuring academic integrity in the era of ChatGPT
Cotton, D., Cotton, P., & Shipway, J. R. (2023) · doi:10.35542/osf.io/mrz8h · Cited by 3,079
Baidoo-Anu, D., & Owusu Ansah, L. (2023) · doi:10.2139/ssrn.4337484 · Cited by 3,004 · cited 1×
“Data produced from ChatGPT may be inaccurate or biased [23], which users need to be aware of to ensure these deficiencies are not propagated in their works.”
Background
Research works printed but not linked: 4
- Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digital Health. 2023;2(2):e0000198. https://doi.org/10.1371/journal.pdig.0000198. · DOI · Paper page
- Shidiq M. The use of artificial intelligence-based chat-gpt and its challenges for the world of education; from the viewpoint of the development of creative writing skills. Proc Int Conf Educ Soc Humanity. 2023;1(1):353–7. ISSN 2986-5832.
- Trust T, Whalen J, Mouza C. Editorial: ChatGPT: Challenges, opportunities, and implications for teacher education. Contemp Issues Technol Teacher Educ. 2023;23(1):1–23.
- Dijkstra R, Genç Z, Kayal S, Kamps J. Reading Comprehension Quiz Generation using Generative Pre-trained Transformers. 2022.
Cited by
Sorted by citations
Items 201–300 of 574
Boon or Bane? Evaluating AI-driven learning assistance in higher education professional coursework
Sun, Y., Yang, H., Yu, H. K., & Suen, R. (2025) · Education and Information Technologies · doi:10.1007/s10639-025-13642-1 · Cited by 12
“…of Large Language Models (LLMs) like ChatGPT (Malfatti, 2025), have dramatically lowered the barrier for individuals, regardless of their technological expertise, to access and utilize the sophisticated capabilities of AI for generating high-quality, context-aware content (Meyer et al., 2023).”
Ratnam, K. K. Y. (2025) · Digital Health · doi:10.1177/20552076251362070 · Cited by 16
Sun, F., Li, L., Meng, S., Teng, X., Payne, T. R., & Craig, P. (2025) · Frontiers in Robotics and AI · doi:10.3389/frobt.2025.1635419 · Cited by 9
Vibe Coding in Practice: Building a Driving Simulator Without Expert Programming Skills
Fortes-Ferreira, M., Alam, M. S., & Bazilinskyy, P. (2025) · International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutomotiveUI) · doi:10.1145/3744335.3758482 · Cited by 9
Henadirage, A., Chathuranga, B. T. K., & Gunarathne, N. (2026) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-026-00578-w · Cited by 9
“This practice is particularly prevalent due to the challenge faced by respondents for whom English is not the native language (Meyer et al., 2023).”
The Rise of ChatGPT and the Demise of Bloom's Taxonomy of Learning Stages
Damaševičius, R. (2023) · Advances in educational technologies and instructional design book series · doi:10.4018/979-8-3693-0205-7.ch006 · Cited by 9
From statistics to deep learning: Using large language models in psychiatric research
Hua, Y., Beam, A., Chibnik, L. B., & Torous, J. (2025) · International Journal of Methods in Psychiatric Research · doi:10.1002/mpr.70007 · Cited by 8
Opening a conversation on responsible environmental data science in the age of large language models
Oliver, R. Y., Chapman, M., Emery, N. C., Gillespie, L., Gownaris, N. J., Leiker, S., Nisi, A. C., Ayers, D., Breckheimer, I., Blondin, H. E., et al. (2024) · Environmental Data Science · doi:10.1017/eds.2024.12 · Cited by 11
“…widespread availability of LLMs will shape how we train the next generation of environmental data scientists, though the advantages and disadvantages of using systems like ChatGPT in the classroom are still being explored (Baidoo-Anu and Owusu Ansah, 2023; Meyer et al., 2023; Milano et al., 2023).”
BackgroundMethodEtori, N. A., & Gini, M. L. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-60606-9_11 · Cited by 10
“The increasing scalability and accessibility of LLMs, as demonstrated by models like ChatGPT [26], have broadened their reach and impact [34].”
Leveraging LLM: Implementing an Advanced AI Chatbot for Healthcare
Mhatre, A., Warhade, S. R., Pawar, O., Kokate, S., Jain, S., & Emmanuel, M. (2024) · International Journal of Innovative Science and Research Technology (IJISRT) · doi:10.38124/ijisrt/ijisrt24may1964 · Cited by 9
Equity Issues Derived from Use of Large Language Models in Education
Bispo, E. L., dos Santos, S. C., & Matos, M. V. A. B. D. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-63235-8_28 · Cited by 8
Augmenting Virtual Labs with Artificial Intelligence for Hybrid Learning
Murali, R., Ravi, N., & Surendran, A. (2024) · IEEE Global Engineering Education Conference (EDUCON) · doi:10.1109/educon60312.2024.10578649 · Cited by 8
“LLM-based chatbots are used to increase student engagement, facilitate group activities, create interactive learning tools, and provide immediate feedback and assessment [8].”
BackgroundEngine of Innovation in Hospital Pharmacy: Applications and Reflections of ChatGPT
Li, X., Guo, H., Li, D., & Zheng, Y. (2024) · Journal of Medical Internet Research · doi:10.2196/51635 · Cited by 12
“Although ChatGPT enhances the efficiency of daily work in hospital pharmacy and provides assistance in scientific research, it still faces some technical hurdles and challenges [28,44-46].”
Yu, B., Xing, H., Ge, W., Zhou, L., Yan, J., & Li, Y.-A. (2024) · PLoS ONE · doi:10.1371/journal.pone.0310724 · Cited by 9
“These models are generally based on deep learning architectures, such as transformers, facilitating impressive performance in diverse NLP tasks [30,31].”
Exploring Explainable AI in Large Language Models: Enhancing Transparency and Trust
H., I., Jo, A. A., & Raj, E. D. (2024) · International Conference on Advances in Computing and Communications (ICACC) · doi:10.1109/icacc63692.2024.10845370 · Cited by 17
Large language model-generated clinical practice guideline for appendicitis
Boyle, A., Huo, B., Sylla, P. A., Calabrese, E., Kumar, S. S., Slater, B. J., Walsh, D. S., & Vosburg, R. W. (2025) · Surgical Endoscopy · doi:10.1007/s00464-025-11723-3 · Cited by 8
Do you actually need an LLM? Rethinking language models for customer reviews analysis
Xiao, Y., Li, Y., Chen, S., Barker, H., & Rad, R. (2025) · Artificial Intelligence Review · doi:10.1007/s10462-025-11308-5 · Cited by 9
Design and Implementation of a General Chemistry Course that Promotes AI Use
Clark, T. M., & Tafini, N. (2025) · Journal of Chemical Education · doi:10.1021/acs.jchemed.5c00728 · Cited by 9
Generative AI and Simulation Modeling: How Should You (Not) Use Large Language Models Like ChatGPTPreprint
Akhavan, A., & Jalali, M. S. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4675409 · Cited by 8
A Review of Using Chatgpt for Scientific Manuscript Writing
Nafea, A. A., AL-Ani, M. M., Khalaf, M. A., & Alsumaidaie, M. S. I. (2024) · Babylonian Journal of Artificial Intelligence · doi:10.58496/bjai/2024/002 · Cited by 10
“Journals like Science and Nature have issued guidelines, stating that AI-generated text or figures are not allowed in published papers and that researchers using LLM tools should document their use in the acknowledgments and methods section[15].”
BackgroundSustainability Integration of Artificial Intelligence into the Software Development Life Cycle
Trinh, E., Funke, M., Lago, P., & Bogner, J. (2024) · IEEE International Conference on Software Architecture Companion (ICSA-C) · doi:10.1109/icsa-c63560.2024.00044 · Cited by 9
“Lastly, we recognize the well-known tendencies of AI to produce plagiarized code or contain bias [56].”
BackgroundEffective prompting with ChatGPT for problem formulation in engineering optimization
Vu, H. N., Wang, K., & Wang, G. G. (2025) · Engineering Optimization · doi:10.1080/0305215x.2025.2450686 · Cited by 7
Bin‐Nashwan, S. A., Sadallah, M., Benlahcene, A., & Ma’aji, M. M. (2025) · foresight · doi:10.1108/fs-04-2024-0089 · Cited by 8
Daraio, C. C., Leo, S. D., & Ferazzoli, F. (2025) · International Transactions in Operational Research · doi:10.1111/itor.70054 · Cited by 8
Oğuz, E. (2025) · Assessing Writing · doi:10.1016/j.asw.2025.100981 · Cited by 12
Bridging LLMs, Education, and Sustainability: Guiding Students in Local Community Initiatives
Jurišević, N., Nikolić, N., Nemś, A., Gordić, D., Rakić, N., Končalović, D., & Kocsis, D. L. (2025) · Sustainability · doi:10.3390/su172210148 · Cited by 7
Pérez-Portabella, A., Arias-Oliva, M., Castillo, G. P., & de Andrés-Sánchez, J. (2025) · Computers and Education Open · doi:10.1016/j.caeo.2025.100323 · Cited by 9
Intertextuality in Pre-service Teachers’ Argumentative Essay in Raising AI: Practices and Beliefs
Martono, Drajatı, N. A., Rochsantiningsih, D., & Wijaya, S. A. (2023) · Register Journal · doi:10.18326/register.v16i2.186-205 · Cited by 9
Söilen, K. S. (2024) · Foresight-Russia · doi:10.17323/2500-2597.2024.2.6.20 · Cited by 10
Boruff, J. T., Kraft, M. A., & Carroll, A. J. (2023) · Journal of the Medical Library Association JMLA · doi:10.5195/jmla.2023.1826 · Cited by 8
“However, the scientific community is far from consensus that this technology is entirely negative, with some researchers already advocating for seeking out opportunities for leveraging LLMs throughout the research life-cycle [9].”
BackgroundAliyeva, A., & Sarı, E. (2023) · Cureus · doi:10.7759/cureus.48366 · Cited by 7
“It effectively eradicates errors that could potentially mar the readability and credibility of a research document, thereby acting as a reliable first line of defense against linguistic imperfections [2].”
BackgroundAI-Powered Test Question Generation in Medical Education: The DailyMed ApproachPreprint
van Uhm, J., van Haelst, M. V., & Jansen, P. R. (2024) · medRxiv · doi:10.1101/2024.11.11.24317087 · Cited by 4
Fluency Illusion: A Review on Influence of ChatGPT in Classroom Settings
Kumar, S., Mikayelyan, A., & Vorfolomeyeva, O. (2026) · Information · doi:10.3390/info17030299 · Cited by 11
Digital Acculturation in the Era of Artificial Intelligence
Durampart, M., Bonfils, P., & Roméro, M. (2024) · Palgrave studies in creativity and culture · doi:10.1007/978-3-031-55272-4_4 · Cited by 6
James, T. G., Guetterman, T. C., & Molina-Azorín, J. F. (2024) · Journal of Mixed Methods Research · doi:10.1177/15586898241287294 · Cited by 6
“Over the past decade, and especially in the past year, the presence of arti fi cial intelligence and other automated tools has signi fi cantly increased in research, and academia is currently in a “ technological evolution ” with the increase in availability of these tools (Meyer et al., 2023).”
Generative AI Assertions in UVM-Based System Verilog Functional Verification
Radu, V., Dranga, D., Dumitrescu, C. M., Tăbîrcă, A. I., & Ștefan, M. C. (2024) · Systems · doi:10.3390/systems12100390 · Cited by 8
Carbon-Aware Workload Shifting for Mitigating Environmental Impact of Generative AI Models
Zhang, E., Wu, D. D., & Boman, J. (2024) · IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData) · doi:10.1109/ithings-greencom-cpscom-smartdata-cybermatics62450.2024.00087 · Cited by 6
Automatic conversational assessment using large language model technology
Bergerhoff, J., Bendler, J. E., Stefanov, S. A., Cavinato, E., Esser, L., Tran, T., & Härmä, A. (2024) · International Conference on Education Technology and Computer · doi:10.1145/3702163.3702169 · Cited by 8
Generative AI in graduate bioprocess engineering exams: Is attention all students need?
Xiao, Z., Lee, E., Yuan, S., Ding, R., & Tang, Y. (2025) · Education for Chemical Engineers · doi:10.1016/j.ece.2025.05.006 · Cited by 6
Sodangi, U., & Isma’il, A. (2025) · Journal of Artificial Intelligence Machine Learning and Neural Network · doi:10.55529/jaimlnn.52.13.23 · Cited by 13
AI in the Writing Process: How Purposeful AI Support Fosters Student Writing
Siddiqui, M. N., Feliciano, V. A., Pea, R. D., & Subramonyam, H. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-98459-4_14 · Cited by 11
ChatGPT as a Virtual Peer: Enhancing Critical Thinking in Flipped Veterinary Anatomy Education
Martín‐Alguacil, N., Avedillo, L., Mota-Blanco, R. A., Marañón-Almendros, M., & Gallego-Agúndez, M. (2025) · International Medical Education · doi:10.3390/ime4030034 · Cited by 10
“In line with the TAM, students who found ChatGPT useful and easy to use were more likely to adopt it; its interface supports engagement and collaborative learning [49].”
The Rise of Large Language Models: Evolution, Applications, and Future Directions
Rahmani, A. M., Hemmati, A., & Abbasi, S. (2025) · Engineering Reports · doi:10.1002/eng2.70368 · Cited by 7
Zhang, B. (2024) · Journal of Contemporary Educational Research · doi:10.26689/jcer.v8i1.5951 · Cited by 6
Chew, Y. J., Yin, O. S., Pang, Y. H., & Lim, Z. Y. (2024) · Forests · doi:10.3390/f15060923 · Cited by 6
“LLMs face several challenges, including hallucinations [17,18], issues with bias and fairness [18–21], data accuracy [18,22,23], data security or leakage [19,22], and ethical concerns [18,20,22].”
BackgroundThe Practice of Enhancing Learning and Scientific Innovative Abilities Using LLM-Based AI Tools
Wang, B., Zhang, X., Li, S., & Wang, Y. (2024) · International Conference on Computer Science and Technologies in Education (CSTE) · doi:10.1109/cste62025.2024.00038 · Cited by 6
Ye, X., Zhai, W., Du, J., Zandt, S. V., & Ye, Y. (2024) · Frontiers of Urban and Rural Planning · doi:10.1007/s44243-024-00046-x · Cited by 6
“The advent of ChatGPT in 2022 signifies a transformative shift in pedagogical and research practices (Gruda, 2024; Meyer et al., 2023; Wang et al., 2024).”
Kaliterna, M., Žuljević, M. F., Ursić, L., Krka, J., & Duplančić, D. (2024) · Scientific Reports · doi:10.1038/s41598-024-77576-3 · Cited by 13
“Since LLMs undoubtedly offer a potential for novel and innovative applications in academia and higher education 16–18 , it is important to explore their potential in generating various types of textual material intended for different purposes.”
Exploring the Role of ChatGPT in Science Education for Asia-Pacific and Beyond: a Systematic Review
Park, H. K., & Martin, S. N. (2024) · Asia-Pacific Science Education · doi:10.1163/23641177-bja10079 · Cited by 9
Wiktor, S., Dorodchi, M., & Wiktor, N. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893324 · Cited by 6
“For instructors, these models can generate coding tuto-rials as well as questions for homework and exams [25] [31].”
InfluentialLarge language models for biomolecular analysis: From methods to applications
Feng, R., Zhang, C., & Zhang, Y. (2024) · TrAC Trends in Analytical Chemistry · doi:10.1016/j.trac.2024.117540 · Cited by 7
Khawaldeh, A. M. (2024) · International Journal for the Semiotics of Law - Revue internationale de Sémiotique juridique · doi:10.1007/s11196-024-10199-z · Cited by 7
Allowing AI co–authors is a disregard for humanization
Bao, A., & Zeng, Y. (2024) · Accountability in Research · doi:10.1080/08989621.2024.2420812 · Cited by 8
“Today’s generative artificial intelligence (AI) are capable of producing text that humans cannot distinguish from human-authored content (Hassanin and Moustafa 2024; Meyer et al. 2023; Teubner et al. 2023).”
Sequential Large Language Model-Based Hyper-Parameter OptimizationPreprint
Mahammadli, K., & Ertekin, Ş. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5362197 · Cited by 6
Tong, S. T., DeTone, A., Frederick, A., & Odebiyi, S. (2025) · Communication Education · doi:10.1080/03634523.2025.2477479 · Cited by 4
“Commentary and scholarship on the use and application of generative AI in universities have focused mainly on issues of integrity, originality, and creativity among students ’ (Chan, 2023b) and faculty researchers ’ written work (Giray, 2024; Meyer et al., 2023).”
Değerli, Y. İ., & Değerli, M. N. Ö. (2024) · Assistive Technology · doi:10.1080/10400435.2024.2416495 · Cited by 6
“ChatGPT interacts with users in a conversational style, employing human-like language to answer questions and create content (Meyer et al., 2023).”
Das, A., Bhattacharya, M., Abdelhameed, A. S., Lee, S., & Chakraborty, C. (2025) · Molecular Therapy — Nucleic Acids · doi:10.1016/j.omtn.2025.102546 · Cited by 5
“15 OpenAI released the GPT-4 model, an MLLM (multimodal LLM) that can be trained with image, audio, or video.”
Jo, H. (2025) · Journal of Organizational Change Management · doi:10.1108/jocm-02-2024-0094 · Cited by 6
Learning Design with an AI Assistant
Divjak, B., Svetec, B., Vondra, P., Bađari, J., & Grabar, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-98414-3_15 · Cited by 5
Flathmann, C., McNeese, N. J., Sengupta, S., & Johnson, E. (2025) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3764591 · Cited by 9
König, L., Zitzmann, S., & Hecht, M. (2025) · Research Synthesis Methods · doi:10.1017/rsm.2025.10053 · Cited by 5
“1,2 Their adoption can expedite the research process, thereby contributing to the ongoing acceleration of scientific output.”
The AI Guilt Complex: Moral Emotions and Ethical Dilemmas in Academic Technology Adoption
Vassallo, D. (2026) · Journal of Academic Ethics · doi:10.1007/s10805-026-09726-3 · Cited by 6
“Large language models can now generate text ‘often indistin - guishable from that written by humans’ (Meyer et al., 2023, p. 1), assist with literature synthesis including formulating Boolean queries for systematic reviews (Wang et al., 2023).”
Akyıldız, S. T. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1783566 · Cited by 7
Predictive typing method for Persian office automation
Nouraei, B., Shanbehzadeh, J., & Asghari, P. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.107792 · Cited by 5
Ilma, A., & Sampurna, R. A. (2024) · Peradaban Journal of Interdisciplinary Educational Research · doi:10.59001/pjier.v2i1.152 · Cited by 8
ChatGPT as a Tool for Equitable Education in Engineering Classes
Ghosh, S. (2024) · ASEE Annual Conference & Exposition · doi:10.18260/1-2--48458 · Cited by 8
ChatGPT in Academia: A Holistic Examination of Student Performance and Future Implications
Dahlan, M. A., Omar, R., & Kamarudin, S. (2024) · Studies in systems, decision and control · doi:10.1007/978-3-031-62102-4_51 · Cited by 5
Sun, J., & Wang, Y. (2024) · Digital studies in language and literature · doi:10.1515/dsll-2024-0014 · Cited by 5
Wong, E. K. C., & Chiu, D. K. W. (2025) · Journal of Librarianship and Information Science · doi:10.1177/09610006251331977 · Cited by 8
“Librarians should educate students on how AI search algorithms and language models can have inherent biases based on their training data and design (Meyer et al., 2023).”
Overview of ChatGPT Model Architecture
Pandey, M. K., & Upadhyay, J. (2023) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-0502-7.ch005 · Cited by 5
Li, L., Dinh, L., Hu, S., & Hemphill, L. (2024) · Research Square · doi:10.21203/rs.3.rs-4887385/v1 · Cited by 13
Gabay, R. A. E., Funa, A. A., & Ricafort, J. D. (2025) · Research Square · doi:10.21203/rs.3.rs-7440784/v1 · Cited by 5
XAI for medicine by ChatGPT code interpreter
Kitamura, K., Irvan, M., & Shigetomi Yamaguchi, R. (2023) · Proceedings of the 2023 5th International Conference on Big-data Service and Intelligent Computation · doi:10.1145/3633624.3633629 · Cited by 5
Lim, B., Seth, I., Cevik, J., Mu, X., Sofiadellis, F., Cuomo, R., & Rozen, W. M. (2025) · Surgeries · doi:10.3390/surgeries6030055 · Cited by 5
Effectiveness of Generative AI Tools in Computer Science and Engineering Education
Waqas, M., Hasan, S., & Ali, A. (2025) · International Journal of Artificial Intelligence in Education · doi:10.1007/s40593-025-00525-2 · Cited by 6
Zhang, J., Xia, Y., Zhong, H., Lu, W., Deng, Q., & Wan, C. (2025) · Mathematics · doi:10.3390/math14010090 · Cited by 4
A Systematic Review and Bibliometric Analysis of Automated Multiple-Choice Question Generation
Mitroulias, D., & Sioutas, S. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10010035 · Cited by 6
Task automation and instructional planning support with large language models: a systematic review
Chontal, G. L., Meléndez-Armenta, R. Á., Degante-Aguilar, E., & Fernandez-Dominguez, F. J. (2026) · Frontiers in Education · doi:10.3389/feduc.2026.1733861 · Cited by 5
Use of large language models in undergraduate medical education a scoping review
Uzoechina, G., & Osajiuba, T. (2026) · Discover Education · doi:10.1007/s44217-026-01239-w · Cited by 5
Engineering Safety and Ethical Challenges in 2045 Artificial Intelligence Singularity
Suo, J., Li, M., Guo, J., & Sun, Y. (2024) · Sustainability · doi:10.3390/su162310337 · Cited by 8
Yousif, N., Youssef, E., & Gad, S. (2024) · European Journal of Education · doi:10.1111/ejed.12889 · Cited by 10
Artificial Intelligence in Oncology
Jeziorski, K., & Olszewski, R. (2024) · Applied Sciences · doi:10.3390/app15010269 · Cited by 4
Navigating Ethical Dilemmas in the Implementation of AI-Driven Educational Technologies
Ismail, M., Madathil, N. T., Alalawi, M., Alalawi, S., & Alrabaee, S. (2025) · IEEE Global Engineering Education Conference (EDUCON) · doi:10.1109/educon62633.2025.11016612 · Cited by 4
“Recent concerns about LLMs like ChatGPT emphasize their potential for content generation but also their risk of false information, requiring human oversight [52].”
Evaluating ChatGPT to Answer Multi-Modal Exercises in Computer Science Education
Ouh, E. L., Tan, K. W., Lo, S. L., & Gan, B. K. S. (2025) · ACM Conference on Innovation and Technology in Computer Science Education (ITiCSE) · doi:10.1145/3724363.3729056 · Cited by 6
Zadeh, P. M., & Sattler, D. (2023) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm58861.2023.10386018 · Cited by 4
“Pre-training on extensive text corpora and subsequent fine-tuning on domain-specific data allow these models to develop a comprehensive understanding of language semantics and syntax [6], [8].”
BackgroundIntelligent Recognition of Tool Wear with Artificial Intelligence Agent
Gao, J., Qiao, H., & Zhang, Y. (2024) · Coatings · doi:10.3390/coatings14070827 · Cited by 4
Ataş, A. H., Cengiz, B. C., & Çelik, B. (2024) · Sakarya University Journal of Education · doi:10.19126/suje.1475474 · Cited by 6
GenAI and Prompt Engineering: A Progressive Framework for Empowering the Workforce
Schuckart, A. (2024) · European Conference on Pattern Languages of Programs · doi:10.1145/3698322.3698348 · Cited by 5
Priya, K., Kamath, A., M, C. K., Praveen, C., N, O. S., & Aaditya, S. J. (2024) · International Conference on Computation System and Information Technology for Sustainable Solutions (CSITSS) · doi:10.1109/csitss64042.2024.10816877 · Cited by 4
Gil, K. L., & Mosquera, E. M. (2025) · Revista Virtual Universidad Católica del Norte · doi:10.35575/rvucn.n74a6 · Cited by 10
AI Generated Code Plagiarism Detection in Computer Science Courses: A Literature Mapping
Simmons, A., de Holanda, M. T., Chamon, C., & Silva, D. D. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893117 · Cited by 6
“Many recent studies have addressed the topic of generative artificial intelligence (AI) and plagiarism in higher education [2]–[7].”
Gürtl, S., Scharf, D., Thrainer, C., Gütl, C., & Steinmaurer, A. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-031-83523-0_35 · Cited by 4
LLMs in Mobile Apps: Practices, Challenges, and Opportunities
Hau, K., Hassan, S., & Zhou, S. (2025) · IEEE/ACM International Conference on Mobile Software Engineering and Systems (MOBILESoft) · doi:10.1109/mobilesoft66462.2025.00008 · Cited by 10
Klasik Türk Edebiyatı Çalışmalarında Yapay Zekâ Kullanımı Üzerine Gelecek Perspektifleri ve Öneriler
Özer, F. (2025) · Akademik Dil ve Edebiyat Dergisi · doi:10.34083/akaded.1705054 · Cited by 4
GenAI as scholarly ally: patterns, pedagogy, and policies in graduate writing research
Гаманюк, В. А., Semerikov, S. O., & Shramko, Y. (2025) · Educational Technology Quarterly · doi:10.55056/etq.965 · Cited by 6
Zhu, S., Wang, Z., Zhuang, Y., Jiang, Y., Guo, M., Zhang, X., & Gao, Z. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4681228 · Cited by 4
Hou, J. (2024) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/d3nc6 · Cited by 30
Khan, N., Daud, H. E., & Shahid, S. (2024) · International Conference on Information and Communication Technologies and Development (ICTD) · doi:10.1145/3700794.3700813 · Cited by 3
“Additionally, concerns regarding overreliance and potential mis-use of LLMs by students highlight the need for ethical guidelines and strategies to promote constructive engagement [8].”
Dartmouth Chat - Deploying an Open-Source LLM Stack at Scale
Stone, S., Crossett, J., Luker, T., Leligdon, L. C., Cowen, W., & Darabos, C. (2025) · Practice and Experience in Advanced Research Computing · doi:10.1145/3708035.3736076 · Cited by 3
Impact of generative models on higher education: exploring opportunities and challenges
Zheng, Q., Yuan, X., & Xu, L. (2025) · Education and Information Technologies · doi:10.1007/s10639-025-13742-y · Cited by 5
“Another study [30] examined the potential benefits and drawbacks of Large Language Models (LLMs) across various fields, including edu - cation in general, specifically programming education and academic writing.”
Metadata
- Authors
- Meyer, J. G., Urbanowicz, R. J., Martin, P. C. N., O’Connor, K., Li, R., Peng, P.-C., Bright, T. J., Tatonetti, N., Won, K. J., Gonzalez-Hernandez, G., & Moore, J. H.
- Affiliations
- Cedars-Sinai Medical Center (US)
- University of Pennsylvania (US)
- Corresponding author
- Meyer, Jesse G.
- Moore, Jason H.
- ORCID
- Venue
- BioData Mining
- Published
- 2023-07-13
- Volume
- 16
- Issue
- 1
- Pages
- 20-20
- ISSN
- 1756-0381 (ISSN-L)
- PMC
- PMC10339472
- OpenAlex
- W4384200891
- Semantic Scholar
- e95b10b9a8415874f24275b665628fc340eee5dc
- Cited by
- 569
- References
- 23
- Citations by year
- 2026: 104 · 2025: 215 · 2024: 220 · 2023: 29
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- BioData Mining (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Institute on Aging (AG066833)
- National Institute of General Medical Sciences (R35GM142502)
- National Institute of Allergy and Infectious Diseases (R01AI164481)
- U.S. National Library of Medicine (R01LM011176)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- large language models, ChatGPT, academic writing, artificial intelligence in education, AI-assisted programming, plagiarism, algorithmic bias, AI accuracy, research tools, ethics of AI use
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
