The integration of artificial intelligence into the healthcare sector, once heralded as a transformative force for efficiency and improved patient care, is now facing intense scrutiny from frontline workers. At HCA Healthcare, the nation’s largest hospital chain, a sophisticated AI-powered scheduling tool named Timpani, developed in collaboration with technology giant Palantir, has become a focal point of discontent among nurses. What was intended to streamline operations and eliminate bias has, for many, devolved into a source of exhaustion, frustration, and concern for patient safety.
For the past four months, Amber Retzloff, a critical care nurse in Florida with a decade of experience, has meticulously documented her interactions with Timpani. She requested to work 50 specific 12-hour shifts, a standard practice for many nurses seeking to manage their personal lives alongside demanding careers. However, the AI system, designed to optimize staffing, has consistently disregarded over half of her preferences. Instead, it has frequently assigned her to disparate shifts, often resulting in back-to-back-to-back working days. This relentless schedule has left Retzloff feeling "mentally drained." While acknowledging that not every request can be accommodated, she states that the AI’s performance in honoring preferences falls far short of the human managers who previously handled scheduling by hand.
Retzloff’s experience is not an isolated incident. It mirrors the sentiments of numerous nurses across HCA Healthcare’s extensive network. Since its rollout began in 2023, Timpani has been implemented at approximately 130 of HCA’s 190 locations. Retzloff’s hospital adopted the system nearly two years ago, providing early insight into its operational impact. "No one is happy with Timpani," Retzloff stated, a sentiment echoed by five other nurses interviewed by WIRED, who also reported widespread dissatisfaction with the software. Public forums, such as Reddit, have also become platforms for purported HCA employees to voice similar grievances about the tool.
The core of the nurses’ allegations centers on the potential for Timpani to compromise patient care. They claim the AI-driven schedules have led to understaffed, imbalanced, and exhausted care teams. The software, operating with what nurses describe as minimal human oversight, routinely schedules insufficient numbers of nurses or fails to deploy enough experienced veteran staff during critical shifts, particularly on Sundays. Furthermore, it frequently disregards nurses’ stated preferences for staggered shifts, working nights and weekends, or securing specific days off. This has forced nurses to dedicate more time and effort to negotiating shift trades or appealing their schedules than was necessary before the AI’s introduction. There is also a perceived increase in absenteeism, with nurses opting to use paid or unpaid time off to avoid undesirable assignments, a move that carries the risk of termination if done too frequently.
Retzloff, who also serves as a union leader with the National Nurses United, recounted recent shifts where she was the sole senior nurse among four junior colleagues. This imbalance necessitated delaying care for critically ill patients to mentor less experienced staff through less complex cases, a situation she attributes directly to the AI’s scheduling decisions. "HCA is letting AI dictate," she asserted. "We are overriding clinical judgment and the human-to-human piece of health care with an app."
Historically, manual scheduling by managers involved direct consultation with nurses, especially when requests deviated from standard assignments. The introduction of Timpani has reportedly shifted this dynamic, leading to shorter notice for alternative arrangements and an increase in temporary reassignments to different teams. A non-union nurse in Texas, who requested anonymity due to fear of retaliation, expressed nostalgia for the pre-Timpani era, stating, "I can’t remember any issues before. Everything came out on time."
HCA’s Defense and Developer’s Role
Harlow Sumerford, an HCA spokesperson, has countered these claims by emphasizing that nursing leaders, not Timpani, hold the ultimate authority in final scheduling decisions. He dismissed assertions that the tool is "designed to reduce staffing at the expense of patient care" as misrepresentative. Sumerford articulated HCA’s objective as ensuring "the right caregivers, with the right skills and experience, are available to meet the needs of our patients while also considering our colleagues’ scheduling preferences." He also indicated that HCA is committed to continuously improving Timpani based on feedback from its nursing staff.
The development of Timpani is rooted in Palantir Foundry, a robust software platform designed for the aggregation and analysis of diverse data streams across an organization’s operations. The system utilizes algorithmic forecasting of patient volumes and anticipated staffing needs to automatically generate nursing schedules. When Timpani was launched in 2023, Michael Schlosser, HCA’s former top innovation executive, projected that it would yield significant savings in time and money, while also eradicating favoritism in shift assignments. In a publication from July of the following year, Schlosser further elaborated that Timpani had demonstrably reduced the monthly hours managers dedicated to scheduling, decreased reliance on expensive contract nurses, and improved staff retention. He attributed this efficiency to the AI’s ability to generate "stronger initial schedules," noting that over 98 percent of these schedules included a "mix of skill and experiences."
HCA’s "responsible AI policy" publicly underscores a commitment to safeguarding individuals from "physical" or "emotional" harm. Company executives have stated that the development of Timpani involved consultations with nurses and an empathetic approach. Marketing materials from HCA have featured nurses commending the tool for its ease of use and time-saving capabilities. However, these positive accounts stand in stark contrast to the experiences reported by many other HCA nurses who feel ignored and disrespected.
Broader Concerns with Palantir Technology
The issues at HCA are not unique to their operations. Rayus Radiology, a major US imaging chain, began experiencing scheduling errors shortly after implementing a similar tool built on Palantir Foundry earlier this year. Internal documents reviewed by WIRED revealed that while the software was designed to populate scheduling fields based on doctor’s orders, it frequently linked requests to incorrect patient profiles, listed tests or scans that did not align with the original orders, or fabricated personal patient details. Schedulers at Rayus suspect that the use of non-standard forms, illegible handwriting, and insufficient human review contributed to these systemic errors.
A spokesperson for Rayus, Kit Crancer, stated that "broad characterizations" do not accurately reflect the company’s "processes, safeguards, or commitment to safe, accurate, and timely patient care." He added that Rayus employs technology "to support, not replace, appropriate human judgment and oversight."
A Palantir spokesperson, who declined to be named, characterized their software as a means of presenting information more effectively and stressed that their clients bear responsibility for the data, policies, and decision-making processes.
The Human Cost: Exhaustion and Anxiety
The implementation of Timpani has coincided with a period of significant growth for Palantir, with its sales nearly doubling to approximately $2 billion per quarter. This surge is driven by an increasing number of organizations seeking its expertise in data analysis and task automation. The healthcare sector, grappling with escalating costs and persistent staffing shortages, has become a particularly eager market. Cleveland Clinic, for instance, has credited Palantir software with enhancing patient capacity by improving the forecasting of ER discharges and replacing manual coordination efforts.
However, Palantir’s expansion has not been without controversy. The company has faced criticism from civil rights organizations regarding its work with defense and immigration agencies. The stated political leanings of its co-founders, Peter Thiel and Alex Karp, have also drawn scrutiny. In the United Kingdom, Palantir secured a substantial $440 million contract with the National Health Service (NHS) aimed at reducing appointment wait times. Nevertheless, nurses and activists have vociferously campaigned for the contract’s cancellation, citing concerns over patient privacy and data security.
At HCA, the sentiment among some nurses is a desire to see Palantir’s involvement removed entirely. While academic studies suggest that AI-based scheduling holds significant promise, nurses involved with Timpani report it has become a considerable source of anxiety. "Everyone checks immediately, and every month you hear, ‘Man, why did it schedule me like that?’" remarked Lee Barker, an HCA nurse in Missouri.
Initially, some local HCA managers possessed the ability to directly edit Timpani’s schedules. However, nurses now report that problematic shifts must be flagged to a centralized team at HCA headquarters in Nashville. This appeals process reportedly occurs without direct communication or the opportunity for nurses to provide in-person explanations. According to Barker and Retzloff, if an appeal is denied, nurses are left to attempt shift swaps, with success rates reportedly hovering around 50 percent.
Barker, who also serves as a National Nurses United leader at an HCA psychiatric center, claims that several colleagues have resigned due to issues with Timpani. Furthermore, they allege a rise in workplace injuries within the first year of the tool’s deployment. The union representing approximately 500 nurses at the psychiatric center has filed a grievance against HCA, arguing that the company is failing to adequately honor scheduling preferences. This matter is reportedly progressing towards arbitration.
In the early stages of Timpani’s implementation, HCA’s Schlosser acknowledged that healthcare workers might be slow to adopt new technologies due to the critical nature of their work, where "they have people’s lives in their hands." He envisioned a future where frontline nurses would actively advocate for more advanced tools, predicting that "then it’ll be fun." However, three years into Timpani’s widespread adoption, many nurses feel a distinct lack of enjoyment and a growing sense of unease.
A Timeline of Implementation and Discontent
- Nearly Two Years Ago: Retzloff’s hospital in Florida implements the Timpani AI scheduling tool.
- 2023: HCA Healthcare begins a wider rollout of Timpani, co-developed with Palantir, across approximately 130 of its 190 locations.
- July (Previous Year): Angelique Russell, a former HCA data science manager, files a lawsuit alleging wrongful termination for raising concerns about Timpani, accusing HCA of data deletion practices that hindered auditing and potentially violated healthcare laws.
- Ongoing: Nurses across various HCA facilities report persistent issues with Timpani, including disregarded shift preferences, back-to-back shifts, and concerns about patient care due to understaffing and exhausted teams.
- Current: National Nurses United, representing a significant portion of HCA’s nursing workforce, continues to advocate for greater transparency and has rejected HCA’s requests for information regarding the technology. Grievances are being filed, and some nurses are considering legal action or seeking employment elsewhere.
The conflict between HCA’s stated goals of efficiency and improved care through AI, and the on-the-ground experiences of its nursing staff, highlights a critical juncture in the adoption of advanced technologies within essential public services. As AI continues its integration into healthcare, the voices of those directly impacted—the nurses and, by extension, the patients—will be crucial in shaping a future where technology truly serves, rather than burdens, the delivery of care. The ongoing debate surrounding Timpani and similar systems underscores the need for robust oversight, transparent development, and a commitment to human-centered decision-making in the pursuit of healthcare innovation.
