As we mark 10 years of Peer Review Week, we sat down for a roundtable with three Frontiers Chief Editors to discuss the reasons why scientists peer review, how open dialogue elevates valid science, and responsible use of AI.
When carried out transparently and collaboratively, peer review is an active partnership that builds valid, trusted science; sharpening methodologies and strengthening the final article.
At Frontiers, safeguarding that trust relies on a three-pronged approach to quality: dedicated research integrity teams, independent editorial boards, and responsible AI and technology.
In a roundtable discussion that took place in our Lausanne headquarters, three of our Chief Editors unpack the value of openness and transparency in peer review, and how trusted science ultimately relies on expert human judgment, even as technological tools advance.
We interviewed:
Professor Pietro Ghezzi, Field Chief Editor of Frontiers in Immunology
Professor Vassiliki Boussiotis, Specialty Chief Editor of Frontiers in Immunology and Frontiers in Cancer Immunity and Immunotherapy
Professor Simon Mitchell, Specialty Chief Editor of Frontiers in Immunology
Moderated by Julie Mo Svalastog, Research Integrity Manager at Frontiers
A researcher’s time is valuable and in demand, so what drives scientists to accept a review assignment when it lands in their inbox? Our editors argue that when researchers set aside time to review, they step into an additional role: not just architects of science, but stewards of its integrity.
Pietro Ghezzi: The reason for doing peer review is a sort of moral imperative, because science needs it.
Vassiliki Boussiotis: Reviewing is contributing to science, by supporting other people’s research and learning from that.
Simon Mitchell: One of the ways we stay up to date with the field is reviewing papers. If we see an interesting abstract, we want to understand what they’re doing, normally just from a place of, ‘Wow, this sounds like cool science, I’d love to understand it.’ So we’ll review it.
The roundtable reflected on how constructive feedback strengthens valid research, maximizing its impact once published.
Vassiliki Boussiotis: I feel that my manuscripts always get better after the review process. A reviewer might point to something critical or trigger new thoughts for me or my team to do additional work and make our eventual manuscript better.
Professor Mitchell: It’s a really good feeling to get a review that says, ‘This would be beneficial to your paper. This extra experiment that you haven’t done would really tie it all together.’ Reviewers really give a lot to your study when that happens.
Frontiers’ interactive online review gives authors and reviewers a direct platform to address questions collaboratively, keeping the focus on sharpening the findings and reaching consensus on sound science.
Pietro Ghezzi: In the Frontiers system, everything happens on a discussion forum where authors and reviewers can discuss live. The key benefit is transparency. That discussion with colleagues mirrors what happens when you go to a conference or have a chat in the corridor.
Vassiliki Boussiotis: When someone knows everybody will read what they write – the authors, the other reviewers, the editors – they are far more professional and instructive.
Pietro Ghezzi: We also publish the names of the reviewers who endorse the paper. That comes down to answerability. They are answerable for the judgment they made.
Vassiliki Boussiotis: It demonstrates professionalism and shows: ‘I am responsible for what I ask, I don’t have bad intentions, and I only want to make this work better with my recommendations.’
In modern publishing, AI performs vital frontline work to protect quality at scale: filtering out low-quality submissions before peer review, flagging ethical red flags, and matching manuscripts with the right expert reviewers. But as discussed by the editors, as these tools take on more operational weight, the boundary between AI support and human, scientific evaluation must remain crystal clear.
Simon Mitchell: If you want to make a judgment call, that still needs a human. I don’t know if AI will ever be able to be critical enough to another human and say, ‘I don’t think that’s good enough’. That’s not something to aim for in peer review. All science must be defensible; you can’t wash your hands of accountability. That’s all that matters.
Pietro Ghezzi: I draw the line on the reviewer’s recommendation to accept or reject. Do we need that to be a permanent line? In a way, yes. Otherwise, you end up with a machine writing a paper, a machine assessing the paper, and a dialogue between machines.
The roundtable closed on the core mission of peer review: treating authors fairly while ensuring every published claim is supported by solid evidence.
Simon Mitchell: I want everyone to come away with a sense of fairness and an understanding of the process. As long as you understand why a decision was reached, and you feel the process was constructive and open, you can take that forward constructively.
Vassiliki Boussiotis: It is vital to communicate clearly what needs to be done for a paper to improve, and to make sure the conclusions and claims are strictly supported by the experimental data. We have to be meticulous, so that every published paper adds one reliable piece of knowledge to the field.
As technology advances, our commitment remains clear: AI will continue to assist and safeguard against misconduct, but the integrity, expertise and critical judgment of our editors and reviewers will always drive the decisions that shape the scientific record.
To learn more about our rigorous review process, explore our peer review guidelines.
September 15, 2026
Editor
Share on WeChat
Scan with WeChat to share this article
Get the latest research updates, subscribe to our newsletter