Despite the growing number of institutional and journal policies that require data‑sharing, most researchers still do not make their data Findable, Accessible, Interoperable and Reusable, resulting in low compliance with FAIR principles (Wilkinson et al.2016). When needing to substantiate that in Data Availability Statements (DAS), authors frequently use the phrase “data available upon request” as a workaround. Although such statements imply willingness to provide access, previous studies have demonstrated that they rarely translate into actual data sharing. Requests are often ignored or declined, and when responses are given, the process may involve obstacles such as restrictive conditions, lengthy negotiations, or ambiguous data‑ownership issues, ultimately rendering access impractical (e.g., Savage & Vickers, 2009). This gap between stated and enacted openness suggests that DAS often serve more as a formal compliance marker than a genuine mechanism for transparency. To evaluate how this practice unfolds in our own institutional context, we analysed a set of publications from our research organisation published in 2024 and found that approximately one quarter relied on the “available upon request” formulation. We subsequently contacted the corresponding authors of these papers and requested either that they deposit their data in an appropriate repository or share their datasets with us directly so that we could assess the quality, completeness, and documentation of the materials. The responses revealed a reluctance to engage in data sharing. A substantial number of authors did not reply, while others declined for reasons such as unclear data ownership, ethical constraints, participant confidentiality, or concerns about the time required to prepare datasets for reuse. Only a minority of researchers were willing to provide access to any materials, and even when data were shared, the level of documentation varied significantly. These findings underscore how the phrase “data available upon request” often masks deeper structural, ethical, and procedural barriers that prevent data from becoming accessible. In addition to reporting these empirical findings, the paper reflects on the ethical challenges associated with studying data that are nominally accessible but effectively hidden behind interpersonal communication and negotiation. Issues arise regarding how to design fair and respectful request procedures, how to interpret non‑responses without penalising researchers acting under legitimate constraints, and how to balance the need for methodological rigor with sensitivity to disciplinary norms and institutional pressures. By examining both the outcomes of data requests and the dilemmas behind stimulating cooperation, we highlight how data availability is not a simple technical attribute, but a socially situated practice shaped by expectations, professional identities, and the everyday realities of research work.
Slavec et al. (Tue,) studied this question.