Constructing a Socio-Legal Framework Proposal for Governing Large Language Model Usage and Application in Education

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a thorough SLR that systematically documents methodology and findings. The section about detection of LLMs is relevant and related to the paper's intention. Final recommendations based on the SLR are comprehensive and logical. This statement in the last paragraph is astute, "Regulation is defined after the emergence of the technology that needs to be regu- 470 lated, and future research could go in the direction of monitoring the compliance of these 471 two processes."

Author Response

Dear reviewer, Thank you for your positive comments. Sincerely, the authors.

Reviewer 2 Report

Comments and Suggestions for Authors

A current work with current and timely references, with an appropriate methodology for selecting sources and offering a clear and useful synthesis of the selected contributions. In short, a work that serves as a framework for other research to delve deeper into the multiple themes it outlines, offering a base scenario to address the normative challenge in relation to education with new educational technologies. It is complex, however, and there are difficulties when it comes to transferring the physical and territorial construct, as well as the temporal one, of citizenship that was born with classical Greece and that developed from Machiavelli and habeas corpus to the Independence of the United States and the French Revolution and that continued to advance after the Second World War with the expansion of the rule of law, but that, faced with globalization, suffers the challenges of trans-territories and with the emergence of the hyperreality of spaces without digital cartography - in this sense, Baudrillard's intuitions are relevant. This new space without physical territory, without clear borders and without time requires a reflection on how to transform the status quo of the Internet citizen, as well as their ways of relating in relation to generative artificial intelligences. We find ourselves at that crossroads between post and transhumanism while the Silicon Valley gurus who encourage us to confine ourselves behind screens create schools for their children without technology or with clear restrictions on the use of the same. If we think that the ethical basis that emanates from politics, since Aristotle, tells us that humans are human because they grow and learn to regulate themselves and to live together, also to develop among humans, based on the shared concept of Paideia, will we be human when we are educated by algorithms? Or will we be moving from zoon politikon to digi zoon. What seems to unite the authors referred to in the work is what Fabris points out, following the philosophical approaches of Vattimo and the Italian postmodern school, the need for netics, an ethics for the new virtual spaces, as a complement to the idea of ​​soft law, which develops in the next generations an autonomous intersubjective morality that enables them to face with possibilities the challenge of living together in the new digital west and to undertake, from ethical criteria, the civilization of the wild digital territory. It is a timely, necessary work, which serves as a space for debate but which requires and we need a greater reflective depth from the anthropological, sociological, philosophical, historical and neurological, also from the pedagogical point of view in order to imagine or think about what kind of humanity we want to project for a future with a greater presence of the non-human and, of course, all this in a neoliberal context where the interests of a few are defended at the expense of the rest of humanity and at the expense of the sustainability of this little blue marble where we are temporarily staying. Congratulations, it has been enriching to read your article.

Author Response

Dear reviewer, Thank you for your positive comments. Sincerely, the authors.

Reviewer 3 Report

Comments and Suggestions for Authors

The article, as far as the systematic literature review methodology is concerned, is correct. However, I do not understand the final conclusions (403-455). If the literature review has been done, these conclusions should not be justified with data directly extracted from that review. It seems to me that the authors make a not very justified leap between what would be the results of a systematic review, which should talk about the data found, and a final reflection of their own on how to summarise or integrate those results but which has been done without justification with the data. I would revise this part to elaborate such a detailed justification.
On the other hand, although I found it very informative and learned a lot, I do not understand the role of section 5 on ai-detection tools. That section needs to be somehow related to the main topic of the article, i.e. how to build legislation on the use of AI in education. I suggest elaborating and justifying this by explaining how the cited authors suggest integrating these tools as part of the regulation to be applied.

Author Response

Dear reviewer, Thank you for your kind recommendations. We are grateful that your comments and recommendations have significantly improved the quality of our paper.

Comment 1. (However, I do not understand the final conclusions (403-455). If the literature review has been done, these conclusions should not be justified with data directly extracted from that review. It seems to me that the authors make a not very justified leap between what would be the results of a systematic review, which should talk about the data found, and a final reflection of their own on how to summarise or integrate those results but which has been done without justification with the data. I would revise this part to elaborate such a detailed justification.)

Respones 1 (We accept your comment related to the request for data justification in our conclusion. Accordingly, in the final part, we have provided information about the authors on the basis of which we came to these recommendations.)

Comment 2. (On the other hand, although I found it very informative and learned a lot, I do not understand the role of section 5 on ai-detection tools. That section needs to be somehow related to the main topic of the article, i.e. how to build legislation on the use of AI in education. I suggest elaborating and justifying this by explaining how the cited authors suggest integrating these tools as part of the regulation to be applied.)

Response 2 (The role of section 5 is to point out the technological aspect of detecting text generated with LLM, which we can conclude is a lost battle in advance, because the authors doubt that it is possible to create a system that reliably detects text generated with LLM. This contributes to recommendations for the affirmative use of LLM in education through the use of a soft law approach. By the way, section 5 was transferred to section 2.2.)

All the mentioned changes are visible in the new version of the text. We have included tracking of changes in the MS Word document.

Sincerely, the authors.