AI is accelerating how research is produced and evaluated, but it cannot decide which questions are worth asking or whom knowledge should serve. At the opening of the 2026 Responsible Research Summit, Prof. Grzegorz Mazurek, Rector of Kozminski University, spoke about the future of research in the age of AI with leaders of organizations that set international standards for responsible research and management education: Dr. David Steingard, Head of PRME; Prof. Alfons Sauquet of EFMD Global; Dr. Lily Bi, President and CEO of AACSB International; and Prof. Agnieszka (Aggie) Chidlow, Chair of RRBM and a scholar at the University of Birmingham. The discussion focused on the limits of automation, the measurement of research impact, and the future of systems for publishing and evaluating research.
Can the current research publishing system survive the rise of AI?
Prof. Grzegorz Mazurek: We talk a great deal about how AI is accelerating academic work. We are also seeing an abundance of content submitted to journals, leaving reviewers and editors with a growing burden as they try to determine whether what they are reading is genuine scholarship. I would therefore like to ask about the entire system through which research is conducted and published. As AI becomes increasingly important, can a model built around journals, publishers and reviewers remain sustainable? Will it undergo a transformation similar to the one the music industry experienced two decades ago? The ecosystem will understandably try to preserve the status quo, but the momentum of innovation may prove stronger. David, how do you see it?
Dr. David Steingard: Let me ground the question in the RRBM Honor Roll, an initiative that recognizes the best scholarship produced by our community. I started the project several years ago, and Andrew Karolyi helped move it forward. Marina Papanastassiou, one of the incoming co-editors-in-chief, is with us today.
The Honor Roll evaluation process is now fully automated. We built a tool called ChatSDG+RR7. It can read a paper, examine it against the Seven Principles of Responsible Research and all 17 Sustainable Development Goals, assign a rating, provide feedback, and even suggest how the work might be applied. For the Honor Roll, this has changed the role of an editorial board of roughly 100 people.
In a paper co-authored with RRBM co-founder David Reibstein and Mark Normandin, we showed that the tool evaluates papers more accurately and with less human bias and error. It works well because it was trained by people who knew what to look for. As long as we keep HIL – the human in the loop – we will be fine.
Dr. Lily Bi: I did not grow up professionally in higher education, so if any of my comments are uncomfortable, please understand that they come from a place of care for this sector. As CEO of AACSB, my job is to be its chief advocate wherever I go.
AI will change research, education, and virtually everything we do. It will not change the two fundamental goals of responsible research that have guided RRBM for 11 years and that AACSB shares: knowledge must be credible and useful. Credibility means academic rigor and an appropriate quality process. Usefulness means addressing a relevant topic and generating impact – or at least beginning with a clear intent to have impact.
What will change is how knowledge is created, disseminated and put to use. AI will revolutionize knowledge creation, whether we like it or not. A literature review that used to take six months can be completed much faster. But the purpose of research is not the literature review itself. Methodology is a means. The end is a credible and useful result. We should not confuse the means with the end. Knowledge creation will become cheaper and may move close to being free. The same will happen to dissemination.
The more fundamental change concerns the actual purpose of research. The current system rewards publication rather than impact, often on the grounds that impact is difficult to measure. We should therefore ask: when does a researcher begin thinking about impact? A conventional route is to identify a gap in the literature: others have studied several adjacent areas but not one particular issue, so filling that gap is expected to generate new knowledge and a publishable paper. There is another way to begin – with a question that is relevant in the real world and with an understanding of the change the research is intended to support.
The intent to have impact should be present when the research question is formulated, not added at the end of the project. That is why the reward system needs to be revolutionized. The volume of publication output alone is no longer an adequate measure of research value.
Prof. Alfons Sauquet: I have more questions than answers. We already know that MIT Sloan Management Review will publish its final issue in September. The rationale is that knowledge will increasingly be disseminated through other media, including podcasts and digital formats. The way knowledge is distributed will certainly change.
I would not, however, dismiss the value of the literature review. There is beauty in that process. Researchers learn the landscape of a field and the questions that mattered in the past, and through that process they gain a better understanding of the major questions of the present. We stand on the shoulders of giants because those giants made the effort required to bring knowledge to where it is today.
Research always contains uncertainty. We begin with a problem, but we do not yet know exactly what kind of knowledge we will need to solve it. AI will offer shortcuts, but some of the substantive frameworks of the research process should remain.
When I think about the future of journals and reviewing, I also ask: who will mentor a young scholar and help that person become a better researcher? I remember a young researcher who later received an award described as the “Nobel Prize of Energy.” She told me that after submitting her first article, she entered into a conversation with a senior scholar who taught her how to refine her questions – and did so in a deeply human way. Speed will undoubtedly change us, but we need to ask how much human contact is required for someone to grow as a researcher and as a person. How do we preserve that part of the process when not everything has merely instrumental value?
There is also beauty in trying to solve a problem before fully understanding either the tools or the possible consequences. Those consequences are not always positive. Michael Jensen offers a useful example. He developed an influential theoretical model of how companies should be managed and disciplined by markets. In practice, that theory contributed to companies being broken up, bought and sold for profit. Jensen himself was surprised by the consequences and spent the later part of his career reframing his theory because he believed it had been distorted.
James March used to say that he liked to speak to his students and young researchers as acolytes gathered around a teacher. I would like us to retain something of that spirit in the age of AI. Artificial intelligence is here to stay; the question is how we shape and use it so that communities of knowledge can survive.
Prof. Agnieszka (Aggie) Chidlow: Since 2015, RRBM – as a scholarly community and a global movement – has asked one question: what is scholarship actually for? Scholarship is one of the core value propositions of higher education institutions, and it helps build trust in those institutions around the world.
RRBM is not questioning the value of rankings as such. We are asking whether ranking systems truly measure the quality and relevance of scholarship. In May, AACSB launched its Global Research Impact Framework in Dublin. A survey conducted as part of the framework’s development found that 87 percent of deans and 82 percent of faculty agree that the definition of impact needs to extend beyond the academic ivory tower to include society, business practice and public policy.
We should therefore ask whether rankings actually measure the value of scholarship. We also know, including through the work of RRBM co-founder Peter McKiernan, that AI poses serious challenges for science. As guardians of research, we must safeguard its credibility and relevance.
The question is no longer whether we use AI. AI is already here, and we cannot simply opt out. The real question is how we use it and how we educate future leaders to use it responsibly. RRBM, its founders and its strategic partners are developing initiatives and frameworks intended to guide that path.
We are not rejecting rankings. We are asking whether they measure what they claim to measure or whether they reinforce a publish or perish culture. Should research instead be guided by the idea of publish to flourish? We should also ask why the research agenda is driven by particular journals rooted in particular contexts. We are meeting in Poland, in the evolving context of Central and Eastern Europe, where language presents an additional challenge. That too belongs in the conversation about the credibility and relevance of scholarship.
Romantics, pragmatists or fatalists?
Prof. Grzegorz Mazurek: In your daily work, you meet many researchers and people who use AI in their professional roles. Let us try to place them in three groups. The first are romantics: they are excited about the future, want to do new things with AI, and approach it optimistically. The second are pragmatists: AI is already here, so they need to use it efficiently and produce good research. The third are fatalists: they fear the end of the world they know, feel that their definition of the researcher is disappearing, and cannot see what their future will look like. Which of these groups currently has the most energy, and which has the least?
Dr. David Steingard: Let us call them romantics, pragmatists and fatalists. We are trying to move toward romantic pragmatism. You should be able to fall in love with a new tool that helps you express your deepest desire to make an impact, while also recognizing that you need structure and a sound system around it.
PRME is currently developing two new journals. One of them is AI and Responsible Management, created with Emerald Publishing. Its mission brings together the use of AI for responsible management, the responsible conduct of research, and research on impact-oriented topics. I have long believed that we need to get ahead of these changes – not to control the narrative, but to give people the tools they need.
At an AACSB conference in Denver in 2024, survey findings were presented on attitudes toward AI in the business school community. I am quoting from memory, but roughly 42 percent of surveyed faculty believed that people should not even go near AI, while around 70 percent had not engaged with it substantively. I guarantee those figures would be different today. AI is advancing extraordinarily quickly – models are emerging that are so powerful that some are withdrawn. Over the past two years, we have already seen movement toward romantic pragmatism. RRBM and the theme of this summit can help take that movement further.
Dr. Lily Bi: Let us take a quick poll of the room. Who considers themselves a romantic? Who is a pragmatist? And who is a fatalist? The result resembles what I usually hear from deans: pragmatists clearly predominate, although we also have a romantic and a fatalist in the room.
Prof. Alfons Sauquet: It is a useful analogy. Deans are usually pragmatists: they move with change rather than trying to fight such a powerful current. AI is one of the great turning points, perhaps comparable to the arrival of the printing press. Before print, a monastery might have had a single book, read aloud every day because only one copy existed. The printing press created a different world, in which people began reading individually and silently. AI may bring change on a comparable scale and open worlds that are difficult for us to imagine today.
We therefore have to be pragmatic. At the same time, I would like us to preserve the spirit of knowledge creation – a process with value and beauty of its own.
Prof. Agnieszka (Aggie) Chidlow: As a scholarly community, we will have to be pragmatic about tools that are developing at an exponential rate. They can support empirical scholarship from a methodological perspective and help us investigate issues that matter to different communities as part of the pathway toward relevance.
Knowledge production, however, happens at the individual level. It is the individual researcher who is embedded in a wider system. We should therefore ask how that system supports the person so that knowledge produced in different contexts is credible and useful. This is the challenge we are trying to navigate across the entire knowledge production system.
We must not forget that the fundamental resource in this system is the researcher, who is exposed to all of these pressures. It is our duty to protect people from excessive pressure. Otherwise, we will destroy the system’s capacity to produce knowledge.
Prof. Alfons Sauquet: It is also worth remembering why we are here. AACSB, EFMD, PRME and RRBM have come together to give real meaning to responsible research. We do not have every answer, but this is exactly the kind of conversation we need. After 11 years of RRBM, we should continue it.
If not points, how should societal impact be measured?
Maria Lipińska, a PhD candidate at Kozminski University: If we move away from points and toward societal impact, someone will probably want to measure that impact with another points system. How should impact be defined and assessed? As university staff, we are continually wrestling with the “ghost of points.” Is qualitative evaluation the alternative? How would you respond to that challenge?
Dr. Lily Bi: Let us begin with questions that come before measurement: whom do you want to influence, and what do you want to change? Only then should we ask how to measure it. When researchers begin a project, they should identify the intended beneficiaries or audiences of impact. Those are the people they should ask whether anything changed and, if so, what. Researchers should also specify the intended outcome. Every study will have a group it concerns and a goal toward which it is working.
I will leave the detailed answer to Eileen McAuliffe, AACSB’s Chief Thought Leadership Officer, who led the development of the Research Impact Framework and will speak about it later in the summit.
Prof. Agnieszka (Aggie) Chidlow: To answer the question, the research process must be designed so that knowledge dissemination and relevant stakeholders are planned from the outset. Impact is not something that occurs at one discrete moment. It develops over time. Researchers need mechanisms that allow them to track how their work and the phenomenon they study influence the intended audience – locally, nationally or globally.
We often fail to do this because the current system does not encourage it. We are trying to change the system so that thinking about impact is embedded in research design from the beginning.
Prof. Alfons Sauquet: It can be understood as a funnel. You start broadly, with climate change, poverty, or another theme that attracts your interest. As the work progresses, you narrow the problem and begin to see clearer connections between the outcomes of the research process and their possible consequences. It is difficult to define impact precisely at the start. You may have a broad understanding of whom the work could affect, and then make it more specific as the project develops – and as you mature as a researcher.
Dr. Lily Bi: Let us look at the issue from one level higher. For decades, even centuries, knowledge creation has focused on discovery. Researchers completed the entire journey in order to make a discovery and create new knowledge. We now have an opportunity to shift the emphasis from discovery to deployment. That is impact.
If the researcher reward system is based primarily on measuring output, it is time to bring that model to an end and build another one. Exactly how to change the system is a difficult question, but the future of research depends on whether we can deploy knowledge in industry, policy and society. That is the true value of what we do.
AI will not change this. It will not replace human beings in defining the problem or formulating the research question. It may replace parts of the literature review process, or accelerate them dramatically, but it will not replace judgment. When AI makes it possible to generate papers at massive scale, someone will still have to evaluate both the process and the result.
We should therefore strengthen the part of the work in which humans create value, rather than compete with AI in areas where the technology has already gained the advantage. The technology industry is moving very quickly, and we do not fully control that pace. But we possess distinctly human strengths. Those strengths should be the foundation on which we build the future of research.
