Social Science in the Age of AI: Unveiling Opportunities, Confronting Biases, and Charting Ethical Pathways

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article deals with a highly topical and important issue. In times of ever-new innovations in the field of AI, it is more than necessary to question the role of the social sciences and to provide ethical guidelines. The article provides a good overview of this in the form of a literature review and also formulates some ethical recommendations.

Only two points should be modified: In section 1.4, the article formulates overly broad research objectives; in particular, it does not provide an interdisciplinary ethical framework, only hints at one, but an article cannot do more than that.

In addition, the healthcare sector is mentioned in a few places (p. 3 and elsewhere), which raises the question of what this has to do with the social sciences, which are the sole focus of the article.

Against this background, I recommend a "minor revision" of this otherwise very exciting literature review article.

Author Response

Please see attached. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses an important and timely topic, but it requires substantial revision to meet the publication's conceptual requirements.

Repetition of Algorithmic Bias Themes

• Section 2.2 (“Social Inequality and Algorithmic Bias”) The manuscript states that AI systems “produce greater error rates” for marginalized groups and that automated hiring and welfare tools “recreate discriminatory patterns.”

• Section 3.1 (“Nature and Sources of Bias”) The manuscript again explains that bias arises from “non‑representative training data,” “proxy variables,” and “institutional environments,” which “reproduce unequal social orders.”

• Section 3.2 (“Fairness as a Multidimensional Concept”) The text repeats that fairness metrics such as demographic parity and equalized odds “may conflict” and that fairness is “context‑dependent.”

The manuscript frequently jumps between education, public policy, and general AI ethics, creating conceptual drift.

For example:

• In Section 2.1, the discussion focuses on communication studies, political science, and predictive models in public policy.

• In Section 2.3, the narrative abruptly shifts to generative AI in education, discussing lesson plans, student writing, and academic homogenization.

• In Section 2.4 and 2.5, the focus remains on automated assessment and university administration, including plagiarism detection and admissions algorithms.

• In Section 3, the manuscript returns to general algorithmic bias, fairness metrics, and governance frameworks.

 

The methodology is not described rigorously or transparently. The criteria for inclusion, scope of literature, and analytical procedures should be explicitly stated, or the claim of systematic review should be removed.

The concept of AI as a Socio-Technical Infrastructure is a good idea; however, the manuscript should clarify what is meant by “infrastructure,” how it shapes epistemic authority, and how it reconfigures social-scientific inquiry.

The manuscript contains grammatical inconsistencies, awkward phrasing, and unclear sentences. A thorough language and style revision is necessary to improve clarity and readability.

Enhance the Conclusions

 

 

Comments on the Quality of English Language

The manuscript requires careful language revision. Several sentences are grammatically incorrect and overly long. Issues include inconsistent verb tenses, incorrect article usage, missing prepositions, and unclear sentence structure. These problems reduce readability and weaken the clarity of the argument.

example : 

" The fast infiltration of the world of artificial intelligence in scientific, economic, and governmental spheres has changed the epistemic and methodological prerequisites of modern social science studies considerably.”

Author Response

Please see the attachment. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

A mapping of the epistemic changes artificial intelligence has brought about in the social sciences is provided in the study "Social Science in the Age of AI: Unveiling Opportunities, Confronting Biases, and Charting Ethical Pathways." Because it reveals AI as a socio-technical infrastructure that reconfigures epistemic authority and the very production of social realities, I myself would embrace the suggested triad - unveiling opportunities, challenging prejudices, and mapping ethical routes. While the appeal to FairAIED and reflexive governance introduces a participatory dimension that echoes abductive ethics - the human must responsibly infer from data rather than passively suffer it - the analysis of algorithmic bias (data, proxy, institutional) deftly captures the "epistemic vice" embedded in black-box models.

However, the article suffers from an overabundance of citations that undermine key synthesis and structural repeats (Sections 4 and 5 overlap). Most importantly, it does not reflect more deeply on the "moral imagination" needed to transform AI from a predictive machine to a reflective partner that can recognize the Global South's epistemic otherness. Overall, the paper shows the way, but it is still too descriptive: ethical suggestions need to go toward a true "reflexive humanism" of the algorithm, or else it would only reinforce the very "algorithmic colonialism" it correctly criticizes. An important but not definitive step forward. The paper is  interesting and publishable, but needs a simplification and a more synthetic treatment

Author Response

Thank you kindly for your comments. We have added the following argument to reflect more deeply on 'Moral imagination' and 'reflexive humanism' in response to the comment suggested by Reviewer 3, where he argues that the article  'does not reflect more deeply on the "moral imagination" needed to transform AI from a predictive machine to a reflective partner that can recognize the Global South's epistemic otherness.'

We have argued in favor of the above comment and added the following paragraph at the end of the Research Objectives section 1.4 in our manuscript:

AI must move beyond merely reproducing statistical patterns derived from dominant datasets and instead develop the capacity to engage ethically with diverse ways of knowing. This is where ethical suggestions must move toward a “reflexive humanism” of the algorithm: not a superficial ethics of fairness layered onto existing systems, but a deeper reorientation in which AI becomes accountable to plural ways of knowing and critically aware of its own embeddedness in structures of domination. In this context, 'moral imagination' is the ability to recognize, respect, and respond to forms of knowledge, experience, and social reality that fall outside hegemonic epistemic frameworks. The emphasis should therefore be on the “epistemic otherness” with knowledge produced in historically marginalized regions that cannot be reduced to deviations from Western norms, but must be understood as grounded in distinct historical, cultural, and political conditions. Thus, our argument calls for AI systems that do not simply reinforce existing hierarchies of knowledge, but rather function as reflective interlocutors capable of acknowledging epistemic plurality and engaging more responsibly with marginalized perspectives.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript has a certain appeal, but it would benefit greatly from a more precise positioning within the theoretical and empirical literature, thereby adding academic value to the paper. Furthermore, the structure could be tightened so that research questions, main arguments, and conclusions are more clearly connected, thereby improving the paper's logical structure and readability. A detailed discussion of the recent research literature would add visibility to the manuscript and situate the research within current academic debates. If relevant, the methodological framing could be clarified, and more explicit justification of interpretive or analytical choices could be provided. Finally, a revision of the language would be necessary to enhance clarity and conciseness, and to ensure the academic quality of the manuscript.

Comments on the Quality of English Language

The English is generally comprehensible, but the manuscript could be improved with careful editing for clarity, grammar, and academic style. Sentences are sometimes very long and contain multiple ideas without transitional phrases, making the argument difficult to follow. In some cases, complex theoretical points are stated in long sentences that would be clearer if broken into several short ones, and there are scattered issues with word choice and phrasing that diminish clarity and obscure the intended meaning. Improved paragraph coherence and clearer topic sentences would also improve the logical flow; therefore, I would recommend a professional language edit or a careful proofreading pass to increase clarity and polish the academic presentation.

Author Response

Thank you for these comments. We agree that clearer signposting of the manuscript’s contribution and argument would improve readability. At the same time, we would note that the paper is already grounded in a defined conceptual framework and engages with relevant scholarship, though perhaps not always in a sufficiently explicit or foregrounded manner. In revising the manuscript, we have therefore focused on making this positioning more visible rather than fundamentally altering the paper’s core argument. We have already clarified the relationship between the research questions, main claims, and conclusions, and made our methodological and analytical choices more explicit. Additionally, we have requested our paper to go through a professional language edit to improve style and cohesion of ideas.