Nurses Blame Palantir Scheduling Software for Errors and Burnout at Hospital Chain
Staff at a major hospital system and radiology network say an AI scheduling tool built by Palantir is producing unsafe shift assignments, adding to a growing labor fight over algorithmic management in healthcare.
A large hospital operator and an affiliated radiology network have deployed scheduling software from Palantir Technologies, and the nurses who use it say the system is generating incorrect shift assignments that disrupt patient care and grind down staff. The complaints, reported by Wired, center on a tool meant to replace manual and legacy scheduling with automated optimization. Instead, workers describe a system that mismatches skills to assignments, creates coverage gaps, and forces last-minute scrambles.
The dispute is the latest flashpoint in healthcare's quiet but accelerating adoption of algorithmic workforce management. Hospitals have turned to vendors like Palantir, which built its reputation on data integration for government and defense clients, to squeeze efficiency out of complex staffing puzzles. Scheduling is an attractive target: nurse labor is the single largest cost line at most hospitals, and chronic shortages make every unfilled shift expensive. But the same optimization logic that trims overtime can also treat a nurse as a generic unit of capacity rather than a clinician with specific competencies and fatigue limits.
Nurses interviewed for the Wired report said the errors are not cosmetic. Assignments that ignore a nurse's unit-specific training or that stack consecutive shifts without adequate rest can put patients at risk, they argue, and the burden of catching those mistakes falls on frontline staff who must then negotiate fixes with managers. The result is what labor researchers call algorithmic stress: workers spend time auditing the software rather than doing the job it was supposed to simplify. Burnout, already at crisis levels in nursing, gets worse.
Palantir has not publicly detailed how its scheduling models are trained, what data they weigh, or how override requests are handled. That opacity is a recurring theme in workplace AI disputes. When a system's output is treated as authoritative, workers who flag problems can be seen as resistant rather than as a safety check. The nurses' complaints echo earlier fights over scheduling algorithms at retailers and logistics firms, where regulators and unions have pushed for transparency and human review. Healthcare adds a sharper edge because the cost of a bad assignment is measured in patient outcomes, not just missed deliveries.
What happens next depends on whether the hospital system treats the nurses' reports as a bug to fix or a grievance to manage. Unions representing nurses have already made algorithmic management a bargaining issue, and state lawmakers have begun eyeing bills that would require human sign-off on AI-generated work schedules in clinical settings. Palantir, for its part, has been expanding its commercial healthcare footprint, and how it responds to these complaints will shape whether hospitals see scheduling AI as a solution or a liability.