Trust, transparency, and disciplinary alignment in research funding evaluations: an illustrative case from the...

The allocation of national AI Centers in June 2025 represents a significant research-policy investment in artificial intelligence in Norway. The broad call, which explicitly included artificial intelligence in the field of education, resulted in 50 applications from large consortia across disciplines and sectors, with an overall rejection rate of 88%. This entailed a substantial aggregate investment of resources across the university, higher education, and research institute sectors, thereby presupposing evaluation processes characterized by a high level of disciplinary expertise, transparency, fairness, and procedural integrity. This article examines the evaluation process associated with the AI Center call, using one application (DLCAIC) as an illustrative case, analyzed through the lens of procedural knowledge, domain-specific expertise, and established research on research evaluation. It explores whether an investment equivalent to approximately half a full-time work year in proposal development can generate meaningful reflection, learning, and improved proposal competence-provided that the evaluation feedback is thorough and well-substantiated, even for unsuccessful applicants. This article is not presented as a conventional empirical study, but as a research-informed Perspective article based on an illustrative case. Its purpose is not to establish generalizable causal claims, but to use a concrete funding-process experience as a point of departure for discussing broader issues of trust, transparency, disciplinary alignment, and procedural fairness in research funding systems. The article further investigates whether committee bias and insufficient disciplinary and sectoral alignment between evaluation panels and application foci may lead to systematic distortions in assessment outcomes. The research literature demonstrates that such distortions may be associated with disciplinary preferences, institutional prestige, cognitive distance, and weakly calibrated evaluation criteria, particularly in interdisciplinary and sector-specific funding schemes. On this basis, the article propose a framework for assessing trust in research funding evaluations based on Panel-domain alignment, Transparency of expert selection, Calibration of self-declared competence, Consistency of written feedback, Actionability/feed-forward value, Procedural symmetry across applications, Disclosure of applicant positionality. The Perspective-article has clear limitations, as it is based on a single call, one application, and one evaluation process. With this caveat, it nevertheless addresses how differing "rules of the game," weaknesses in panel appointment, competence profiles, and the application of evaluation criteria may undermine trust in the research funding system. The Perspective-article concludes by arguing for broader disciplinary representation in evaluation panels, clearer justifications for expert selection, and more transparent evaluation procedures as key measures to strengthen legitimacy and trust in the national research funding system.