Harnessing generative AI to drive responsible business research and accelerate social impact Open Access

Responsible business research has become a key pillar of the broader “business as a force for good” movement, reflecting the growing view that academic work should move beyond theory to address real-world societal challenges (McPhail et al., 2024). Leading this shift is the Responsible Research in Business and Management network (RRBM, 2025) which promotes business research that is both academically rigorous and socially impactful (Jack, 2025). This aligns with wider calls to reshape business academia to confront global challenges and advance the United Nations (UN) sustainable development goals (SDGs) (Eustachio et al., 2024). Recent work underscores the need for business strategies that are not only responsive but also responsible and proactive in driving sustainable global change, urging business research to help meet pressing societal needs (Cornuel et al., 2023).

Despite increasing momentum behind responsible business research (Barrington and Karolyi, 2023), efforts to identify and evaluate such work remain fragmented and inconsistent, limiting the visibility and scaling of impactful scholarship (Steingard and Rodenburg, 2025. Traditional peer review, with its narrow focus on academic rigor, often fails to capture whether a paper meaningfully focuses on positive social impact. In response, the RRBM network launched the Responsible Research Honor Roll in 2022 (RRBM Honor Roll, 2025) to recognize business school scholarship that contributes to the public good.

As Steingard and Rodenburg (2025) observe:

This study centers on the RRBM Honor Roll and applies a custom-built, fit-for-purpose generative artificial intelligence (GenAI) tool – ChatSDG+RR7 (Shinn, 2024) – developed from prior work on assessing SDG alignment (Garwood et al., 2020; Steingard et al., 2022). Used effectively, GenAI can support peer reviewers in making decisions that are more accurate, less biased, better aligned with established standards and more efficiently rendered (Fiorillo and Mehta, 2024). One of the continuing problems with the publishing of academic work is how long it takes work to pass through the review process. With the careful use of AI we are able to provide reviews in a much more expeditious manner. The task of peer reviewers – here assisted by ChatSDG+RR7 – is to identify papers that both address the SDGs and adhere to the Seven Principles of Responsible Research; this constitutes the “ground truth ^[1]” to be discerned.

Despite these advancements, several critical questions remain regarding the reliability, objectivity and epistemic legitimacy of AI-assisted peer review. To this end, our analysis is guided by the following research question:

Our definition of responsible research aligns with AACSB’s view of social impact as activities that create “meaningful, discernible change for the betterment of people, economies, and the environment” (AACSB, 2023, p. 2). Research forms the foundation of how business schools influence real-world practice. When it emphasizes responsibility, it can shape the strategies businesses adopt. Elevating responsible research strengthens the connection between scholarship and practice and encourages organizations to pursue actions that benefit both society and the planet. Shifting from traditional scholarship toward a “new normal” of disruption and impact (Bridoux et al., 2024), responsible business research deepens academia’s engagement with societal needs. Grounded in global challenges (Howard-Grenville and Spengler, 2022), including the SDGs, our approach shows how business research can guide action and deliver measurable social impact (Financial Times, 2024).

We focus on the RRBM network, a key authority in the field of responsible business research. Our analysis centers on the RRBM Responsible Research Honor Roll (Honor Roll, 2025), an editorially curated collection of academic articles recognized as “responsible” by expert reviewers selected by the RRBM network which owns and manages the platform. Established in 2022, the Honor Roll currently includes 364 awardees across 99 unique journals. Evaluations follow RRBM’s normative standards based on the SDGs and the Seven Principles of Responsible Research (Principles of Responsible Research, 2015).

Table 1 presents an approach designed to provide an understanding of responsible business research. A fundamental aspect of responsible business research is its foundational focus: To whom is the research addressed and what type of impact is envisioned? Can researchers “systematically extend or enlarge their research agenda or projects to amplify their impact on the challenges societies face” (Wickert et al., 2020, p. 297)? Notably, the rise of responsible business research reflects a shift from scholar-to-scholar oriented research to research that engages broader societal concerns, scholar-to-society (Hoffman, 2021).

At the core of this discussion is the need to expand – rather than discard – the traditional view of relevance in evaluating academic research quality. Research should be considered “high quality” (Jack, 2025) when it demonstrates both rigor and relevance. Rigor refers to the methodological precision and thoroughness that ensure reliability and validity which together form the foundation of meaningful scholarly progress. Relevance, by contrast, is more contested – raising critical questions:

The pervasive “publish or perish” mentality in academia (Steingard and Rodenburg, 2023) often results in an excessive focus on publication quantity and citation counts. This emphasis can overshadow the importance of researching and highlighting work concerned with social impact. Scholar-to-society relevance is frequently sidelined because paper counts and citations are easy to quantify, while social impact is more challenging to measure. This paper introduces a novel evaluative AI technique grounded in the core principles of responsible research (Tijdink et al., 2021) and applied to the domain of business scholarship.

As responsible business research gains momentum, it is essential to develop a robust theoretical and applied framework to ensure its impact is both meaningful and measurable. Peer review plays a central gatekeeping role in determining what qualifies as scholarly value. Traditionally, it prioritizes methodological rigor, contextual fit and strong theoretical grounding – criteria aligned with norms of doctoral training. As illustrated in Figure 1, the relevance or societal responsibility of research is often filtered through this evaluative lens (Desmond, 2024). Consequently, judgments about relevance may be sidelined or inconsistently applied during peer review. This paper builds on that insight by proposing a renewed view: that academic rigor and real-world relevance must operate in tandem to advance responsible research that contributes to social impact.

Given that the goal is to achieve dual relevance – both academic and real-world – there is a need to address the current imbalance. Traditional human-led peer review has consistently favored academic relevance over societal impact. This highlights the need for tools that can support reviewers in overcoming this entrenched bias (Tyser et al., 2024). By learning to emphasize real-world relevance, peer review can better align with responsible business research goals. This is why we turn to AI assistance to introduce a non-human input that could offer more objective contributions to the peer review process. We introduce AI assistance as a complementary, non-human input that can enhance objectivity and consistency in the evaluation process. At the same time, we are mindful of the early and sometimes contentious acceptance of AI in peer review. Accordingly, we use AI assistance solely as an editorial aid within a human-in-the-loop (HIL) framework (Natarajan et al., 2025), with the editor reserving the discretion to recommend a full human peer review in cases of uncertainty. This ensures that human judgment remains the ultimate arbiter of scholarly quality and impact.

To support human reviewers in evaluating responsible business research, we apply AI-assisted peer review (Kousha and Thelwall, 2023; Zhu et al., 2026). Using GenAI, we codify the principles of responsible management research and embed them into the peer review process through our study of the RRBM Honor Roll. The goal of using AI assistance in the review process is to enhance peer review’s ability to identify responsible business research through a more principled, systematic, reproducible, and unbiased framework. This strengthens academic work’s capacity to rigorously and intentionally address societal and sustainability challenges. The approach reinforces the theoretical and empirical foundations of responsible research and helps rebalance academic and real-world relevance. By using GenAI, we aim to more effectively identify, exemplify and distinguish responsible business research. We now turn to our conceptual and empirical application of a custom GenAI peer review model for the Honor Roll.

The objective of the RRBM Honor Roll is to recognize published academic work that addresses societal issues. We focus on the RRBM responsible research adjudication process which leads to a designation on the Honor Roll. We use AI-assisted peer review to support the Honor Roll’s evaluation process. This method offers a robust way to identify responsible business research that meaningfully addresses social impact. To begin, it is important to distinguish how peer review is conceived within the Honor Roll. Evolving from Figure 1, Figure 2 offers some slight modifications and operationalizations to this end.

The Honor Roll journal articles reviewed were already published and were presumed to meet minimum academic quality standards. However, recognizing that peer-reviewed status does not always guarantee legitimacy, we implemented an additional layer of quality assurance specifically to verify the credibility of the journals themselves. In light of the global rise of fake and predatory journals, we cross-referenced all Honor Roll submissions with Cabells Journalytics and Predatory Reports databases (Walters, 2022) to confirm their legitimacy. None of the journals represented in the Honor Roll were flagged as predatory. This step was essential as accepting submissions from illegitimate sources would undermine the credibility of a platform focused on responsible business research.

Since the articles are already published, the decision centers on whether to recognize them with Honor Roll selection. This streamlines the review process by eliminating reviewer-author exchanges and revise-and-resubmit cycles. The growing success of AI-assisted peer review in journals and conferences (Tyser et al., 2024) validates the strength of this approach. If AI can handle complex reviews, it is well equipped for a binary yes/no decision, making it ideal for applying Honor Roll standards to submissions.

The AI peer review assistant, ChatSDG+RR7 (Shinn, 2024) was conceived, trained and deployed using the ChatGPT-4 platform and integrated into the RRBM Honor Roll peer review process. We began by analyzing the history of prior Honor Roll decisions. This historical analysis covered a data set of 141 of which had previously been accepted. We then applied ChatSDG+RR7 to evaluate these same papers.

ChatSDG+RR7 evaluated all of these papers and generated basic summary statistics about the data set. Based on this data, we applied a normal distribution curve, with one standard deviation, to create three categories: recommendations to cut, recommendations to pass and recommendations for human review. These categories were split into 25%, 50% and 25%, respectively. The original criteria for the RRBM Honor Roll were based on two principles encoded into ChatSDG+RR7: Service to Society – the development of knowledge likely to benefit business and society for the ultimate purpose of creating a better world – and Impact on Stakeholders – research that contributes to better business and a better world. Service to Society is interpreted to mean research that addresses societal issues and in this case the UN SDG’s. Impact on Stakeholders is more a measure of the degree/magnitude of the research impact.

Since these standards are fixed, they can be encoded into the ChatSDG+RR7 system. This allows the generation of consistent, standards-based assessments, both qualitative and quantitative. Importantly, the system is designed to provide advisory evaluations, with human reviewers retaining final decision-making authority. By calibrating ChatSDG+RR7 to the precise standards of the SDGs and the seven principles of responsible research the system delivers rigorous assessments that align with RRBM’s vision of responsible management research grounded in both “rigor and relevance” (Tsui, 2019, p. 167).

To evaluate articles for inclusion in the RRBM Honor Roll, ChatSDG+RR7 applies natural language processing to assess full-text academic PDFs against a structured rubric. This rubric is based on the SDGs, the seven principles of responsible research and a training data set composed of previously selected Honor Roll recipients chosen by human editorial board members. The tool analyzes each submission for its relevance to societal challenges, methodological rigor and alignment with responsible research values. Custom scripts and fine-tuning were incorporated to improve accuracy, along with human validation checks and anti-hallucination safeguards.

Once a submission is processed, ChatSDG+RR7 produces both qualitative and quantitative outputs used in the editorial review process. The key component of the output is a qualitative narrative, including a brief article summary (distinct from the original abstract), the top three SDGs the article addresses and an explanation of how each SDG is engaged. These outputs are reviewed by the Honor Roll editor-in-chief who makes the final inclusion decision. This AI-human collaboration ensures that automated machine evaluation is guided by expert human oversight, combining efficiency with principled judgment (see Web Appendix 1: Step-by-Step Submission ExampleLink to the cited article for a detailed description of the entire AI-assisted evaluation and editorial decision-making process).

This study included both previously accepted and new submissions to the RRBM Honor Roll, enabling a more comprehensive analysis of acceptance and rejection decisions. For each article, ChatSDG+RR7 generated a narrative summary explaining how the submission addressed each of the Seven Principles of Responsible Research. This qualitative evaluation was then converted into a quantitative score ranging from 0 to 5, mirroring the scoring method used by human academic reviewers. A separate score was also generated for SDG alignment. These two scores – adherence to the Seven Principles and SDG alignment – were weighted, combined and averaged to calculate a final overall score for each submission.

ChatSDG+RR7 delivered a comprehensive, consistent analysis of how each Honor Roll submission aligned with the established criteria. We now turn to the findings, where we evaluate the performance of both the editor-in-chief and ChatSDG+RR7 and examine how AI-assisted peer review brought AI–human collaboration to life in the RRBM Honor Roll study.

This empirical study on the RRBM Honor Roll is designed to provide insights into the effectiveness of AI–human collaboration, particularly within the subdomain of AI-assisted peer review. How effectively did AI contribute to the Honor Roll reviewing process? We examine key findings from the study, focusing on the collaboration between the human reviewer (editor-in-chief) and ChatSDG+RR7 in evaluating Honor Roll submissions.

Table 2 offers a summary of collaborative decision-making between the editor-in-chief and ChatSDG+RR7 across 322 RRBM Honor Roll submissions. It displays how editorial decisions aligned with AI-generated recommendations, capturing two core metrics: the ai-human evaluation agreement rate and the editor-in-chief submission acceptance rate. Each submission is categorized into one of three AI-generated recommendation tiers: “Pass” (recommended for acceptance), “Cut” (recommended for rejection) and “No Recommendation” (a middle range within one standard deviation, designed to allow greater human discretion).

The findings are categorized into two primary types: convergent assessment and divergent assessment. Convergence refers to cases where both the editor and ChatSDG+RR7 agreed on whether to include or exclude a submission from the Honor Roll, while divergence indicates disagreement between their evaluations. To assess alignment, we used F1 scores, which are performance metrics commonly used in information retrieval and machine learning to evaluate the balance between precision and recall. Precision measures how many of the predicted positives are actually correct, while recall measures how many of the actual positives were correctly predicted. The F1 score is the harmonic mean of the two measures.

These scores do not indicate statistical significance or hypothesis testing; rather, they reflect the practical accuracy of the AI’s decisions compared to human judgments. For a detailed explanation of the threshold-setting logic, based on previously published Honor Roll papers we used in test benchmarking, see Web Appendix 2: Setting Thresholds for Decision-MakingLink to the cited article.

The findings indicate strong alignment between the AI tool and human editorial judgment. When ChatSDG+RR7 recommended a “Pass,” the editor-in-chief agreed in 84.52% of cases (131 of 155 submissions). When the AI recommended a “Cut,” the editor concurred in 100% of cases (13 of 13). No agreement rate was calculated for submissions that received “No Recommendation” designations, which fall within the zone of discretionary human review. Overall, Table 2 demonstrates how ChatSDG+RR7 can meaningfully support and reinforce consistent editorial decision-making.

The editor-in-chief ultimately accepted 197 out of 322 submissions, resulting in an acceptance rate of 61.18%. This rate aligns with historical data from previous Honor Roll evaluations conducted without ChatSDG+RR7 indicating that the integration of AI did not substantially alter editorial judgment. Rather, it enhanced the efficiency of the Honor Roll review process and validated human editorial judgment. As demonstrated in Table 2, ChatSDG+RR7 supported editorial decision-making by offering a check on human inconsistency and bias, aligning with established standards for responsible research and strengthening the rationale and defensibility of scientific rigor in peer review. The demonstrated efficacy of ChatSDG+RR7 provides encouraging evidence for the broader, ethical and effective deployment of AI-assisted peer review systems. An expanded discussion of these findings is available in Web Appendix 3: Extended Discussion of ResultsLink to the cited article and Web Appendix 4: F1 Score AnalysisLink to the cited article.

While there was substantial convergence between the editor-in-chief and ChatSDG+RR7 on acceptance and rejection decisions, this does not fully resolve the challenge of establishing a reliable ground truth (Kang, 2023; a more in-depth exploration of ground truth can be found in Appendix 5: Extended Discussion of Ground Truth in AI-Assisted Peer ReviewLink to the cited article). Is the best judgment being made and by what standards is it justified? It is both illogical and unwise to fully cede human judgment to AI-assisted peer review for two reasons. First, although AI can achieve high levels of accuracy with large data sets, those data sets are often flawed because they are based on human judgment that is inherently inconsistent and biased. When bias is present (Checco et al., 2021), AI is only as reliable as its inputs. Second, this risk is heightened with GenAI, which is prone to errors, biases and hallucinations. Unlike algorithmic AI (Kalmykov and Kalmykov, 2024) which follows fixed rules, GenAI learns patterns to generate new content – making it more vulnerable to inaccuracies (Maleki et al., 2024).

Our findings reveal that AI-assisted peer review reduced human errors and biases, offering a closer approximation to the ground truth. Our analysis suggests that relying solely on human expert consensus may not be sufficient, as it remains prone to inconsistency and bias – especially when compared to AI-assisted evaluation. As detailed in Web Appendix 2: Setting Thresholds for Decision-MakingLink to the cited article, out of 141 previously accepted Honor Roll papers we used in test benchmarking, 18 (12.76%) were rated below the “cut” threshold (<4.1) by ChatSDG+RR7. This level of divergence indicates that nearly one in eight papers endorsed by human reviewers would not have met the more rigorous, evidence-based thresholds of ChatSDG+RR7. In this sense, ChatSDG+RR7 does not merely replicate human judgment but provides a stricter and more consistent evaluation standard – helping to prevent “false positives” that can slip through human-only review. It aligns with growing literature showing that AI-assisted peer review improves reliability, reduces bias and strengthens editorial integrity (Farber, 2024).

One of the unique contributions of our study is that it applies AI-assisted peer review specifically to responsible business research, where societal impact and SDG alignment are central to evaluation. This focus has not been examined in prior large-scale studies of AI-assisted review. In addition, our work embeds the UN SDGs and the seven principles of responsible research into the evaluation process – operationalizing normative standards rarely tested in peer review – and demonstrates these innovations empirically through the RRBM Honor Roll, a curated platform of socially impactful scholarship that represents a real-world application.

This paper set out to examine how GenAI can strengthen the evaluation of responsible business research by identifying work with high potential for social impact in business. Specifically, our research question asked:

Our findings indicate that GenAI, when thoughtfully integrated through a HIL model, enhances the quality of peer review by increasing consistency, reducing bias, speeding up the review process and aligning editorial decisions more closely with normative responsible research standards. These results suggest that AI-assisted peer review is not simply a technical augmentation but a transformative catalyst that advances both the rigor and societal relevance of business scholarship.

This section highlights how ChatSDG+RR7 contributes theoretically to responsible business research by improving the quality and epistemic legitimacy (Kang, 2023) of scholarly evaluation. Too often when a paper is reviewed the specific assessment may depend on the most recent quality of the paper(s) previous reviewed which set a benchmark or even the specific review team selected. Just note the variance in reviews from within the review panel. This study demonstrates that AI-assisted peer review enhances evaluation quality by increasing consistency, reducing bias and on top of that improving efficiency. By applying standardized criteria grounded in the SDGs and the Seven Principles of Responsible Research, the AI system supported editorial decisions that were more consistently aligned with responsible research standards.

First, AI–human collaboration enhances consistency and reduces bias by combining the efficiency of algorithmic assessment with the ethical oversight of human judgment. This partnership maintains methodological rigor while advancing editorial goals focused on social impact in business. Through this synthesis, ChatSDG+RR7 operationalizes a hybrid model of review that balances computational precision with the normative and contextual discernment unique to human evaluators.

Second, the interaction between AI and human reviewers helps clarify for perspective contributors the foundations of research quality. AI-supported scoring and justification provide a transparent rationale for decisions, which strengthens confidence in editorial outcomes and reinforces the epistemic legitimacy of the peer review process. This theoretical contribution underscores that responsible research evaluation is not only a procedural act but also an epistemological one, grounded in shared accountability between human and machine.

Third, ChatSDG+RR7 strengthens the identification of socially impactful research by systematically aligning scholarly evaluation with the SDGs. While alignment alone does not guarantee impact, the tool’s precision and evidence-based justification increase the likelihood that research addressing societal challenges will influence policy, practice and future scholarship. This integration of AI with responsible research principles demonstrates how evaluation systems can serve as vehicles for advancing rigor, relevance and social value simultaneously.

Collectively, these findings advance our understanding of how AI-assisted systems can operationalize responsible research principles. Within the RRBM Honor Roll, ChatSDG+RR7 offers a replicable model for uniting rigor and relevance in scholarly review while helping identify and elevate research with strong societal impact. The tool addresses key flaws in traditional peer review, including bias, inconsistency and the exclusion of diverse perspectives (Checco et al., 2021; Desmond, 2024).

Grounded in the real-time deployment of AI-assisted peer review to the RRBM Honor Roll, our findings are applicable not only here but also generalizable to other academic assessment processes with responsible research standards. The practical implications are substantial: AI-assisted peer review can transform how responsible business research is evaluated, recognized and adopted. The success of ChatSDG+RR7 shows how AI can help human reviewers make more consistent, principled and evidence-based evaluations. This has broader relevance, suggesting AI-driven approaches could be successfully adopted across disciplines.

This study offers a detailed and robust exploration in AI-human collaboration in the subdomain of AI-peer review. It contributes to research by empirically testing AI-assisted evaluation and supporting the theoretical view that AI-assisted review outperforms human-only review. In addition, our “convergent and divergent assessment” framework provides a useful method to define, operationalize and test AI-assisted peer review.

The use of ChatSDG+RR7 in the RRBM Honor Roll presents key research implications for evaluating responsible business. This AI-assisted model integrates SDG alignment with the seven principles of responsible research. Together, these criteria establish a robust benchmark for enhancing objectivity and minimizing human bias in scholarly evaluation. Its standardized criteria improve fairness and transparency – longstanding challenges in traditional peer review. It also encourages broader academic adoption, where responsible research principles guide assessments of contributions to global challenges.

Our contribution aligns with recent evidence that few studies explicitly measure societal impact and that AI can play a vital role in closing this gap (Steingard and Rodenburg, 2025). It also directly addresses the challenge identified by Gohr et al. (2025), who note that while AI is increasingly used in sustainability forecasting, it rarely incorporates deep SDG contextual knowledge. By operationalizing both the SDGs and the Seven Principles of Responsible Research within a peer-review framework, our study offers a concrete way to address these gaps and provides an empirically tested approach for editors seeking to evaluate responsible business scholarship with greater rigor and impact.

Our findings contribute to the growing literature on AI-assisted peer review by extending its application to the domain of responsible business research. Prior studies have demonstrated that AI can improve efficiency and reduce bias in traditional peer review (e.g. Fiorillo and Mehta, 2024; Farber, 2024), yet little attention has been given to how such tools might help editors address current challenges of assessing social impact and producing higher-impact scholarship. This study responds to recent calls for more systematic, rigorous and transparent approaches to evaluating responsible research (George et al., 2021). Our AI-assisted peer review tool helps bridge this urgent gap in relevant literatures.

The theoretical foundation backed up by the reliable results in this study enables transformative social impact research to be shared more rapidly, increasing its real-world application. The RRBM Honor Roll becomes a living repository of responsible scholarship, ready for integration into practice and policy by amplifying high-impact work through its open access global platform (RRBM Honor Roll, 2025).

The RRBM Honor Roll exemplifies the kind of multi-channel dissemination platform that York et al. (2025) identify as essential for translating academic research into real-world impact. By curating and openly sharing responsible business scholarship through an accessible digital interface, the Honor Roll extends research visibility well beyond traditional journal audiences. It creates a living, interactive repository that connects scholars, managers and policymakers. Through open-access dissemination and cross-sector engagement, it operationalizes RRBM’s mission to make responsible business research both academically rigorous and practically actionable. In this way, the platform transforms evidence-based insight into measurable societal benefit realized through the work of practitioners.

This broad reach enables more effective application of academic research to real-world challenges. For business leaders and managers, access to research that has been rigorously vetted through responsible-research criteria provides credible, evidence-based insights that can guide sustainability strategies, stakeholder engagement and organizational decision-making. For policymakers, the availability of AI-assisted evaluations linked to the UN SDGs helps align public programs and regulatory frameworks with empirically supported models of social and economic impact.

By connecting peer-reviewed research directly to implementation contexts, this framework supports “business as a force for good” in tangible ways – encouraging companies and institutions to adopt practices that are both profitable and socially responsible. In doing so, AI-assisted peer review expands the practical utility of academic knowledge and reinforces the reciprocal relationship between research, management and policy innovation.

AI-assisted peer review warrants continued investigation into how AI–human collaboration is transforming editorial roles, decision-making and accountability across the academic publishing ecosystem. The rapid rise of AI in research evaluation introduces both promise and risk. Without sufficient governance, unsanctioned or poorly regulated applications can amplify bias, validate low-quality research or legitimize flawed methodologies – particularly in predatory journals that lack robust editorial standards (Checco et al., 2021). These concerns underscore the need for stronger ethical and regulatory frameworks to ensure that AI supports, rather than undermines, the goals of responsible business research.

Beyond governance, Fiorillo and Mehta (2024) identify persistent challenges in over-reliance on automated assessments, incomplete evaluations and the neglect of nuanced academic reasoning. Editors who depend too heavily on algorithmic outputs may overlook scholarly depth and complexity, while bias within AI models can disadvantage underrepresented authors and research domains. Maintaining the integrity of peer review requires balanced human–AI collaboration and rigorous oversight – principles demonstrated throughout this study.

As AI continues to evolve, it is increasingly plausible that AI-assisted peer review systems will become reliable enough for human reviewers to act primarily as final decision-makers, endorsing AI recommendations while maintaining ethical oversight (Heaven, 2018; Kousha and Thelwall, 2023). This development highlights a critical opportunity for future research: to examine how AI–human collaboration in peer evaluation can redefine academic quality by reuniting rigor and relevance. Such inquiry could clarify the evolving meaning of excellence, particularly regarding social impact in responsible business scholarship.

Emerging innovations already address several of these challenges. Markhasin (2025) introduced persistent workflow prompting to reduce subjectivity and improve consistency through standardized evaluation workflows – a principle reflected in ChatSDG+RR7’s implementation. Tools such as the PolicyProfiles Tracker (Sage Policy Profiles, 2025) can further trace the real-world influence of research in policy and practice. Expanding open-access dissemination also remains vital for enhancing the visibility and usability of responsible research across academic, corporate and public sectors.

Overall, addressing these limitations and pursuing these directions can advance the ethical, methodological and practical integration of AI-assisted peer review. In doing so, future work can strengthen transparency, inclusion and societal relevance – ensuring that systems like ChatSDG+RR7 contribute meaningfully to responsible business research and its global impact.

The supplementary material for this article can be found online.