Skip Content
You are currently on the new version of our website. Access the old version .
MDPIMDPI
by
  • Pilar Sánchez-García1,
  • Alba Diez-Gracia1 and
  • Pedro Jerónimo2,*
  • et al.

Reviewer 1: Anonymous Reviewer 2: Anonymous Reviewer 3: Tigere Paidamoyo Muringa

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. The article is well-written with a professional and academic tone.  Key terms such as "generative AI" and "analytical AI " are defined clearly. 

    The language is precise, and technical terminology is appropriate for the target audience.

2. There are minor typographical errors, such as "regulatations" (should be "regulations") and "forthe" (should be "for the") in the abstract. Proofreading will help polish these issues.

    Some sentences are lengthy and dense, which may reduce readability. For example, in the Introduction: "The academic debate on the subject moves between the optimism of its contributions to journalistic workflows and the limitations or risks of its use in the news industry." Could be simplified to: "Academic debates highlight both the benefits of AI for journalism and its associated risks."

3. The abstract could be more concise, focusing on key findings without repeating background information. 

4. The Introduction provides a thorough overview of AI in journalism but could benefit from a sharper focus on the research gap. 

5. The methodology is robust, clearly explaining the snowball sampling technique and coding process.

6. The paper highlights the limitations of generative AI but does not deeply explore its potential benefits or long-term implications for journalism. The role of emerging AI technologies (e.g., ChatGPT, DALL-E) in reshaping the newsroom could be explored further, given their relevance to the study's timeline.

7. The article examines 44 stylebooks, guidelines, and codes of ethics from various media outlets and organizations. While it thoroughly explores ethical self-regulation in journalism, it does not comprehensively address all known ethical codes on AI like UNESCO's AI Ethics Recommendations or IEEE’s Global Initiative on Ethics of Autonomous and Intelligent Systems. Other widely referenced principles, such as Google's AI Principles or the OECD Principles on Artificial Intelligence, are also absent. The author(s) should include global initiatives like UNESCO’s AI ethics recommendations, OECD guidelines, or IEEE's ethical frameworks.

 

Author Response

Please see attach. 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Peer review 24.1.2025

 

Media self-regulation in the use of AI: delimitation of multimodal generative content and ethical commitments to transparency and verification

 

I would like to begin by expressing my gratitude for the opportunity to serve as a peer reviewer and to engage with this manuscript.

 

The paper examines media self-regulation in the use of AI by analysing an international sample (N=44) of editorial stylebooks and internal self-regulatory guidelines published between 2023 and 2024. As put in the abstract, “the research studies how different media outlets, media groups and institutions are beginning to create internal regulations for the use of AI, both from a technical and an ethical perspective.”

 

The study has three objectives (p. 3): “1) Collect and review different editorial style manuals and/or internal guidelines that include recommendations and regulations for the use of AI in international media (local, regional and/or national), and gather their metadata in open access. 2) Analyse the permitted, limited and/or prohibited uses of AI in journalistic production and especially around the generation of multimodal content. 3) Identify the ethical stance and role of human oversight in the application of AI, transparency of usage and data verification.”

 

The topic of the article is highly interesting, important, and timely, aligning well with the scope of Journalism and Media. Despite its compelling focus on media self-regulation in the use of AI, I must respectfully recommend major revisions. In the following paragraphs, I will outline the reasons for this decision, provide critical observations, and offer suggestions to enhance the paper so it better meets the standards of scientific research articles.

 

The article’s introduction and theoretical background sections are still quite brief and should be deepened. AI is utilised in journalism and media organisations for a wide variety of tasks, and these could be described more extensively in the introduction/theory (not just, for example, comment moderation or translations, but also tasks like gathering and analysing background materials, writing, targeting and distributing news, etc.). Additionally, the history and background of media self-regulation could be elaborated on—for instance, what media self-regulation has been like in the past, how it has been implemented, in what contexts, and in what ways, and why? Also, the section on AI ethics is currently rather short. In other words, the author(s) could more thoroughly demonstrate their familiarity with prior research in the field and anchor their study more strongly in its context.

 

On p. 3, it is stated that the article includes hypotheses. Hypotheses might be better suited for quantitative research, where they are proven “true” or “false” using quantitative research methods. Could these “hypotheses” be more accurately described as “background assumptions” rather than testable hypotheses per se?

 

An international dataset has been collected for the study, which is an impressive and ambitious goal. The data collection process is described (p. 4), and the dataset is listed in Table 2. However, it remains somewhat unclear to the reader how exactly these guidelines/documents were identified and selected for the study, and what they represent within the international context? So, what does the dataset tell us about the topic—does it provide a generalisable view of the current global situation?

 

The dataset lacks internal guidelines from media organisations that are not publicly available online but strongly influence operations. How has this been considered, and what are the authors' reflections on this?

 

Although data has been collected globally from different countries, some of the guidelines seem to be (?) still local. What does “internationality” mean here, and how is it reflected in the results?

 

The distinction between “generative AI” and “analytical AI” should be further clarified and justified. It is also worth noting that generative AI can perform analytical tasks, and analytical AI can generate outputs.

 

It seems that the analysis is somewhat incomplete, and the results could be further elaborated. At present, they feel somewhat like a collection of tables. While the most important tables can remain in the main text, longer ones might be better suited as appendices. Also, the article could benefit from adding an evaluation of the study and suggestions for future research at the end.

 

Smaller comments:

  • Alphabetise the keywords.
  • Section 1.1. requires a counterpart, such as section 1.2., that is, there should not be standalone subsections in an article.
  • The reference list should be formatted according to the guidelines of Journalism and Media.

 

 

Despite the major revisions I have proposed, I see a lot of potential in this paper, and I sincerely encourage the authors to continue developing this manuscript. I hope that the feedback provided will assist in refining the work into a publishable article. Good luck!

 

Author Response

Please see attach. 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Abstract

·      The abstract need to be reworked to meet the standard of a structured abstract. There is a tendency to use very long sentences by the authors that affects the meaning of the paper

·      The context of the study can be briefly introduced in the abstract

·      A statement of the implication or contribution of the study can be added into the abstract

Introduction

·      There is a tendency to use very long sentences. For instance, line 28-31 is one statement that 4 lines long. This really affects the readability of the text

·      There is need for a clear discussion of the problem statement. Its not really clear, what really is the problem that the paper seeks to address

1.1.                Self-regulation and ethics of Artificial Intelligence usage in the news media

·      What is needed in this section is for the authors to use short sentences that not difficult to read through

·      Line 105, its not clear who is being cited. There is only (2023, np) without authors name

·      While the theoretical grounding is robust, further integration of studies from the Global South would provide a more balanced global perspective. Incorporate more diverse case studies, especially from regions outside Europe and North America, to strengthen global applicability.

Research Design and Methods

·      Clarify why snowball sampling was chosen over systematic sampling and discuss its limitations more explicitly. Explain the rationale behind choosing a limited sample (N=44) and discuss potential biases introduced by snowball sampling.

Coherence of Arguments and Findings

·      Expand on how findings can be generalized across different media ecosystems.

·      Ensure that all tables (e.g., Table 3) are consistently formatted and adequately explained in the narrative.

·      Ensure tables are adequately described in the text. For example, Table 3 requires clearer linkage to the findings

·      Avoid dense academic jargon to enhance accessibility for broader audiences.

Adequacy of References

·      Add references for datasets and further literature on regulatory frameworks in non-European contexts.

·      There is need to update old reference or support them with recent citations

Conclusions

  • Highlight actionable recommendations for policymakers and media practitioners to address identified gaps.
  • Expand on practical implications for journalists and editors, addressing how they can implement the ethical guidelines discussed.

Overall Recommendation

·       Accept after minor revisions

·       Rationale:

·       The manuscript is well-researched and presents a significant contribution to the scholarship on AI ethics in journalism. While the content is robust and original, minor revisions are necessary to enhance clarity, global inclusivity, and the organization of results and tables. These improvements would elevate the manuscript's accessibility and impact without requiring substantial rework.

 

 

Comments on the Quality of English Language

The quality of english is good. but authors need to do editing and use short sentences. 

Author Response

Please see attach. 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Peer review 11.2.2025

The authors have made all the suggested corrections to the article and provided further justification for their choices. The article has been improved and can now be accepted by the editor for publication in Journalism and Media. Well done and good luck!