A team science blueprint for nurse-led artificial intelligence research: Lessons from building AI tools to detect stigmatizing language in obstetric records.

BACKGROUND: Nurse scientists are increasingly expected to lead research applying artificial intelligence (AI) to advance health equity. However, there is little published guidance on how to structure and lead the multidisciplinary teams required to meet such work demands. PURPOSE: To describe a replicable team science model for nurse-led AI research, derived from a completed three-year project using AI text analysis to detect stigmatizing language in obstetric electronic health records and its associations with maternal health outcomes. METHODS: We present a case-based analysis of a multidisciplinary project (2022-2025) integrating nursing, data science, epidemiology, obstetrics, and informatics. We describe team structure, workflow, and decision-making across five methodological phases. FINDINGS: Four principles for effective nurse-led AI team science emerged: deliberate role design positioning nurse scientists as intellectual leaders; investment in cross-disciplinary translation infrastructure; operationalization of equity at every stage; and realistic resource and timeline planning. DISCUSSION: This blueprint offers practical guidance for nurse scientists seeking to lead technically complex, equity-focused AI research, arguing that team structure, not technical sophistication, is the most critical determinant of success.