In contemporary digital societies, datification increasingly prioritizes algorithmic efficiency over lived human experience. This epistemological shift is particularly visible in medical environments where artificial intelligence (AI) is embedded in diagnostic workflows. This paper examines this transformation through a paradigmatic microcosm: fetal Magnetic Resonance Imaging (MRI), where extreme technological sophistication intersects with profound existential vulnerability.

Here, the pregnant body is translated into data and workflows, widening the interpretive gap between what machines process and what patients experience as fear and ambiguity. Current AI systems often operate as opaque “black boxes.” While deep learning models optimize image reconstruction, this opacity risks silencing the narrative and relational dimensions of care, reducing patients to objects of technical interpretation.

Situated at the intersection of Digital Humanities and medicine, this paper introduces the conceptual framework of Vulnerability-Aware Artificial Intelligence (VAAF). Grounded in feminist ethics, the Ethics of Care, and digital hermeneutics, VAAF proposes a shift toward a “glass box” model. Rather than autonomous decision-makers, AI systems are reframed as supportive infrastructures. By automating technically demanding tasks, AI alleviates cognitive burdens, restoring time for listening and empathetic mediation.

Crucially, ethical and clinical interpretation must remain strictly non-delegable (human-in-the-loop). This contribution argues that healthcare constitutes a critical site for Digital Humanities inquiry, revealing how digital systems shape meaning-making practices. Ultimately, the most human-centered technologies do not maximize automation, but amplify human capacities for interpretation, care, and relational presence.