Image source: Frontiers Media SA
Publishing policy needs to catch up with reality and provide guidance and guardrails, says Elena Vicario
Last week, the country’s largest research funder, the German Research Foundation (DFG), It has announced that artificial intelligence can be used responsibly to review funding applications. Smart move. I lead the research integrity team at Frontiers, one of the world’s largest open access publishers, and my question is, if funders are doing this, why aren’t they publishing? We need more proactive action on AI across publishing, research and policy-making institutions. AI is present and happening throughout the academic publishing lifecycle. As publishers, we can either get ahead of the adoption of this technology or be left behind.
The question is no longer whether AI is part of the research lifecycle. When it comes to peer review in publishing, what is urgently needed now is a policy that lags behind the actual use by authors and reviewers, and the explicit wishes of authors and reviewers. our According to research and a white paper published last year, most reviewers now use AI tools to support their work. Recently, Frontiers’ pilot integrating an AI chat assistant directly into our review platform has reinforced this demand. Initial results from the pilot show that out of more than 10,000 respondents, 80% of users found the tool very useful or helpful. The majority of the remaining 20% want more AI capabilities, not less. Those involved in peer review work don’t want publishers to rein in their use of AI. They want us to help publishers use AI better and more responsibly.
