Registered nurses, university workers, and community activists converged in front of the North Carolina General Assembly in Raleigh on Friday morning to demand what Governor Roy Cooper’s successor has so far declined to provide: a hard stop on unchecked AI expansion in healthcare settings.

The July 24 press conference was organized by the National Nurses Organizing Committee/National Nurses United (NNOC/NNU), the American Association of University Professors NC, UE Local 150 (a workers’ union representing public sector and health care employees), and the North Carolina Data Center Network. The event was timed to respond directly to the AI Leadership Council’s Statewide AI Strategic Roadmap, which Governor Josh Stein unveiled July 1 as a framework for how North Carolina will deploy and regulate artificial intelligence across state government and key economic sectors.

What Nurses Are Demanding

NNU’s position is specific. The union is not asking the state to ban AI in healthcare — it is asking for three things the roadmap does not currently require:

  • A moratorium on new data centers until adequate environmental and worker-impact review processes are in place
  • Enforceable override rights — a legal standard establishing that healthcare professionals and patients, not AI systems, make final care decisions
  • Mandatory disclosure when AI is influencing clinical recommendations, discharge timing, staffing levels, or other patient-affecting decisions
“Nurses are already seeing AI outputs influence care decisions in ways patients don’t know about and clinicians can’t always override. North Carolina has an opportunity to set a national standard for patient-centered AI governance. This roadmap doesn’t do that.”— NNOC/NNU statement, July 24, 2026

What the North Carolina AI Roadmap Actually Says

Governor Stein’s AI Strategic Roadmap, developed by the North Carolina AI Leadership Council, is framed around three goals: protecting residents and their data, preparing the workforce for AI-augmented jobs, and strengthening the state’s economy through AI adoption. The document calls for transparency, fairness, and human oversight as principles, but critics note it does not create enforceable rules or establish a regulatory body with authority to penalize AI deployments that harm workers or patients.

The roadmap addresses healthcare AI as one use case among many — including education, transportation, and public safety — without dedicated sector-specific requirements. It encourages “responsible AI adoption” without defining what would render an AI deployment irresponsible enough to trigger regulatory action. NNU and UE Local 150 argue this is the gap: the roadmap gives good-faith language without teeth.

Why Nurses Are in This Fight

NNU has been tracking AI deployment in healthcare settings for two years and identifies several specific vectors of nurse-patient impact that fall outside existing regulatory frameworks:

  • AI-driven discharge recommendations that flag patients as ready for discharge based on clinical parameter algorithms, sometimes overriding bedside nurse judgment about patient readiness
  • AI-assisted staffing tools used by hospital management to determine shift staffing levels, which NNU argues systematically undercount patient acuity in ways that cannot be effectively challenged in real time
  • Ambient AI documentation that captures clinical conversations and auto-populates the medical record, raising questions about nurse liability when the AI transcription differs from what the nurse intended to document
  • Predictive sepsis and deterioration alerts that carry false-positive rates high enough to generate significant alarm fatigue, affecting how nurses triage and respond to real emergencies

None of these use cases involve AI making clinical decisions autonomously. But each creates situations where AI outputs influence clinical workflow in ways that nurses cannot always correct without documentation burden or pushback from hospital administration. That’s the governance gap NNU wants codified — not a ban on the technology, but a clear chain of accountability when AI-influenced decisions produce adverse outcomes.

The clinical reality

NNU’s position on AI isn’t anti-technology. The union has long supported nurse-facing tools that reduce documentation burden and surface early warning signals. What they’re flagging is a gap in accountability: when an AI-generated discharge recommendation leads to a readmission, who is responsible? The roadmap doesn’t answer that question. Neither does anything in federal law right now. That’s the argument, and it’s a substantive one.

Data Center Moratorium: The Less Obvious Nursing Issue

The demand for a data center moratorium sounds like it belongs to a different debate, but NNU’s argument connects it to nursing workforce conditions. Large AI data centers require significant water for cooling — in North Carolina, a state where drought conditions have affected rural communities, some proposed data center projects draw from water systems that serve hospitals and long-term care facilities in the same region. UE Local 150 has documented cases where rural health system infrastructure competes with data center expansion for grid capacity in eastern North Carolina counties.

The health system connection is indirect but real: rural hospitals that lose reliable power or water infrastructure face operational constraints that ultimately translate to nursing staffing and patient care decisions. NNU is framing data center growth as a health system resilience issue, not purely a labor or environmental one.

National Pattern

North Carolina is not the only state where nursing unions are pushing back on AI governance gaps. NNU affiliates have filed bargaining demands related to AI transparency in California and New York, and the union’s 225,000-member organization has made AI worker protections a stated national priority alongside safe staffing ratios. The Montefiore Health System in New York announced layoffs of clinical support staff in July 2026 citing AI-driven workflow automation — a concrete example of the employment impact NNU is pointing to as evidence that “AI readiness” frameworks need to include workforce protection teeth, not just aspiration language.

North Carolina’s specific profile makes it a meaningful battleground: the state is both a major healthcare employer (UNC Health, Duke Health, Atrium, Novant) and an AI industry growth target, with significant data center investment in the Research Triangle and Piedmont regions. What the state decides in the next legislative session about enforceable AI guardrails will influence both sectors.