Researchers conducted semi-structured, in-depth interviews with 26 ophthalmic nurses selected through maximum-variation sampling. Participants worked in tertiary, secondary and primary eye-care settings across eastern, central and western China, providing coverage of different levels and locations of care. The interviews used the Critical Incident Technique, while the transcripts were examined with Braun and Clarke’s reflexive thematic analysis. Reporting followed the Consolidated Criteria for Reporting Qualitative Research, or COREQ.

Three themes emerged from the analysis. The first, described by the researchers as a “tentative approach,” captured a progression from curiosity to a conditional appraisal of generative AI’s competence. Nurses did not treat the technology as uniformly reliable or unreliable; instead, their trust was calibrated in a graded, task-specific manner. The researchers tentatively called this pattern “gradient calibration,” indicating that acceptance varied with the task and context rather than representing a single position toward generative AI.

The second theme, “guarding professional boundaries,” concerned the anchoring of trust in ethical obligation and the nurse–patient relationship. This focus is particularly relevant to ophthalmic nursing, where visual outcomes can be irreversible, ophthalmic medication requires precision, and visually impaired patients may depend on nurse-mediated support. The findings indicate that the possible reconfiguration of professional roles and relationships by generative AI forms part of nurses’ trust negotiations. They also show that trust is not only a judgment about technical performance: ethical tension and responsibility toward patients influence how nurses engage with the tools.

The third theme placed trust negotiation within the wider organizational ecology of ophthalmic care. Institutional conditions, professional expectations and the workplace context shaped how nurses approached and used generative AI. Taken together, the themes show trust emerging through negotiation rather than as a fixed attitude of acceptance or refusal. The study does not measure whether generative AI improves diagnostic accuracy, patient outcomes or nursing performance, nor does it establish how common particular responses are among ophthalmic nurses. Because the evidence comes from interviews with 26 participants, its conclusions describe the experiences and reasoning of this sample. Transferability beyond the sampled Chinese eye-care settings requires further study.

Subject of Research: Trust negotiation in generative AI use among ophthalmic nurses in China

Article Title: Navigating the unknown: a qualitative descriptive study of ophthalmic nurses’ trust negotiation with generative artificial intelligence in China

Article References: Zhu, M., Zhang, X., Zhu, Y., Zhang, L., Li, R., Li, A., & Qu, L. (2026). Navigating the unknown: a qualitative descriptive study of ophthalmic nurses’ trust negotiation with generative artificial intelligence in China. BMC Nursing. https://doi.org/10.1186/s12912-026-05437-7

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05437-7

Keywords: ophthalmic nursing, generative artificial intelligence, trust calibration, qualitative interviews, nurse–patient relationship, clinical ethics, organizational context, eye care, thematic analysis, Navigating, unknown, qualitative