“Never let a good crisis go to waste,” Churchill is alleged to have said. Scholarly publishing could learn a thing or two from this perspective. For decades, scholarly publishing endured an arguably “self-made” crisis driven by exponential price increases. The solution offered to libraries (buy fixed bundles and get a discount) only deepened the crisis, shaping how academic publishing is perceived by the research community and fuelling an “I can fix it myself” mentality as Diamond OA publishing suggests. To be clear: I fully support Diamond, but wonder whether it is equally suitable for all disciplines.
From my perspective, the debate became stuck on fairly traditional distribution issues while we lost sight of the bigger picture our industry is responsible for addressing: validating and making information available to everyone. OA is still the best answer we have to the serious societal problems like climate change, exhaustion of resources, and health crises we need to find solutions for. Nothing works better than unrestricted access and the freedom to research. And to those who say, “it is too expensive,” I respond: “No, it isn’t! Full OA is the best, most affordable, and most trustworthy way to open scientific knowledge!” Making articles free to read is only the first step. What follows is the unfinished business of transforming OA into Open Science. This we need to embrace because the rapid rise of AI presents formidable challenges and opportunities – from safeguarding the integrity of scholarly content and combating misinformation, to rethinking workflows and reimagining value creation.
Without open licenses and open data, transparency remains selective and accountability incomplete. Ensuring the correctness of research data is especially urgent in an era where machines cannot validate themselves. Human oversight, supported by openly accessible datasets, is essential. Our responsibility now is to take the next step by enabling FAIR data – findable, accessible, interoperable, and reusable – because none of the grand challenges captured in the SDGs will be solved within disciplinary silos. They demand cross-accessibility of data that we are still far from achieving. And if research is to remain relevant in an AI-powered world, we must also ensure that the data we produce are machine-readable and AI-ready, or we risk building an open science framework that is open in name only.
While I describe challenges, I return to my opening statement: within these challenges lies an unprecedented opportunity. Our industry is uniquely equipped with experts who understand academic rigour and quality. We possess deep know-how built over decades, and hold the key to designing efficient, trustworthy AI-powered workflows. Most importantly, the industry can build sustainable financial models that invest not just in innovation, but in long-term stewardship of knowledge. The task ahead is clear: to harness AI not as a threat but as a catalyst – transforming disruption into renewal and ensuring that academic publishing continues to be a cornerstone of credible, accessible, and impactful research in the AI age. This is not just survival; it is a chance to lead, adapt, and thrive.
This article was first published in the Publishing Partners Guide 2026, which was distributed with the January / February 2026 issue of InPublishing magazine. You can register to receive InPublishing magazine here.
