Written By: Ben Rafal
The American healthcare system is at the forefront of modern innovation, yet it remains under pressure. National healthcare spending reached $5.3 trillion in 2024, which significantly exceeds the international average and that of other high-income nations (Center for Medicare & Medicaid Services, 2023). Despite this, the system is burdened by ongoing labor constraints and complex bureaucratic work that does not directly improve patient outcomes. There is an opportunity for artificial intelligence products and other software services that reduce inefficiencies and render healthcare easier to provide. AI tools that save time are an essential short-term adoption trend that will transform healthcare cost structures.
The United States spends significantly more per person on healthcare than other highly developed nations like Germany, Japan, and the United Kingdom, yet the US population has a lower average life expectancy and a higher mortality rate from treatable diseases than each of those countries (OECD, 2025). The binding constraint within the system is not medical knowledge, but rather the scarcity of time and attention paid to sick individuals. Provider effectiveness is measured by both profit margins and the number of minutes dedicated to each patient. This leads to an exceedingly difficult dual mandate and a cycle of staff turnover, inefficient time allocation, and poor-quality care.
Persistent staffing shortages and high rates of burnout among nurses and doctors are critical challenges facing the US healthcare system. The Association of American Medical Colleges projects a shortage of up to 86,000 physicians by 2036, up from about 37,000 in 2021 (AAMC, 2024). Approximately 41% of nurses report intending to retire or switch careers within the next five years, which translates to up to 1.6 million individuals (NCSBN, 2024). This maintains high wage inflation, continually limits the efficiency of the clinical care process, and reduces providers’ average capacity, pressuring both costs and revenue.
Healthcare spending is under intense scrutiny due to rising hospital supply expenses per patient outpacing general inflation between 2019 and 2022 (American Hospital Association, 2023; Bureau of Labor Statistics, n.d.). A separate problem is that the system spends a lot of time and money on highly automatable administrative assignments rather than actual patient care. Carrying out prior authorizations with insurance providers is a concrete example of this burden that could become significantly more streamlined. To contextualize this issue, the most costly part of an MRI may not be the procedure itself, but the required process of billing and documentation. A physician survey from the American Medical Association reports that physicians and healthcare staff spend 13 hours per week completing paperwork and cite this task as a frequent cause of burnout (AMA, 2025).
The short-term objective is to remove the administrative time sinks that stagnate the system. The long run presents more tectonic opportunities for technology to change the roles of doctors and nurses. While still in their early stages, AI-enabled digital health tools are the current focus of widespread adoption. The funding environment for healthcare technology tightened significantly due to higher interest rates and more conservative investor behavior, and has reopened selectively following demonstrated, measurable returns on investment from AI-native applications. AI garnered 54% of digital health funding in 2025, up from 37% in 2024, with capital appearing to be concentrated in the products perceived as most effective, rather than broadly distributed across a wide variety of ventures (Zweig et al., 2026).
Although concentrated in the back office departments of the healthcare system, these tools reduce the administrative time associated with insurance claims. Agentic AI agents can authorize insurance, help schedule follow-up appointments, and suggest next steps in treatment alongside a physician during patient consultations. Waystar has developed software that improves how hospitals and clinics verify insurance coverage, respond to denials, and collect patient payments, helping prevent revenue loss and collection discrepancies, which becomes especially detrimental when provider margins are compressed (Waystar, n.d.). A Microsoft tool records and automatically converts patient conversations into structured notes, so that doctors can focus on evaluating the patient’s case and developing a tailored treatment plan (Microsoft, n.d.). Administrative automation tools pay for themselves so quickly because they can translate the messy reality of healthcare into formats that insurance companies and compliance systems require.
In addition to supporting administrative tasks, AI products are evolving beyond static information hubs and can now serve as dynamic collaborators with healthcare providers in imaging, disease detection, and early warning. The rise of multimodal AI systems that integrate multiple data sources, including MRIs, CT scans, electronic health records, and genomics, provide richer diagnostic insights. A Swedish study found that AI-supported breast cancer screening achieved a detection rate similar to the manual approach while reducing radiologists’ workload by 44% (Lång et al., 2023). In emergency radiology, triage software can flag CT scans that may indicate a pulmonary embolism and help prioritize them so that the most urgent cases are read first (Lamb, 2023). IDx-DR can detect diabetic retinopathy from retinal photos of adults without requiring specialists to review every image (FDA, 2018).
Early warning systems are increasingly prevalent in hospital workflows, aiming to flag patients whose conditions may be deteriorating. However, the widely deployed Epic Sepsis Model, which analyzes over 80 real-time variables, performed substantially worse than expected, underscoring the ever-present risk inherent in AI prediction models and the continued room for improvement with tools that are already immensely powerful (Wong et al., 2021). As of February 2026, at least 1,350 AI-enabled medical devices have been authorized by the Food and Drug Administration, supporting the shift towards AI as a trusted collaborator in healthcare despite the ongoing risks (Dowdell et al., 2026).
Another emerging initiative is a larger shift to lower-cost settings, such as “hospital at home” programs, which have exploded in popularity following the COVID-19 pandemic. Delivering hospital-level care in the home for selected patients, along with remote monitoring and in-person check-ins, reduces the need for inpatient beds and enables the scarce supply of hospital staff to care for more patients. In hospitals permitted to use this model, the 30-day patient readmission rate decreased by 70%, and 30-day post-discharge Medicare spending was 22% lower on average (Levine et al., 2019; CMS, 2024). In-home hospital care will turn “capacity” from a fixed number of beds into a growing network, eliminating space constraints.
When physicians and staff spend hours each week on documentation, and hospitals operate under tight budget constraints, the most successful products are those that give time back and measurably protect a provider’s cash flow. Over the long run, clinical AI will matter most when it demonstrably improves outcomes in real emergencies. The time and money these tools save will offset the pressure from rising equipment costs and low Medicare reimbursements, allowing nurses to focus more time on the patient-facing tasks that require specialized skills and judgment. The next era of healthcare services will be led by products that make the system less bureaucratic and more dedicated to true patient care.
References
2024 AMA Prior Authorization Physician Survey. (2025). American Medical Association. https://www.ama-assn.org/system/files/prior-authorization-survey.pdf,
Center for Medicare & Medicaid Services. (2024, September). Report on the Study of the Acute Hospital Care at Home Initiative. United States Department of Health and Human Services. https://calhospital.org/wp-content/uploads/2024/10/AHCAH_Study_092724.pdf.
De Novo Classification Request for IDx-DR. (2018). U.S. Food & Drug Administration. https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN180001.pdf.
Dowdell, J., Stecklow, S., Terhune, C., & Levy, R. (2026, February 9). As AI Enters the Operating Room, Reports Arise of Botched Surgeries and Misidentified Body Parts. Reuters. https://www.reuters.com/investigations/ai-enters-operating-room-reports-arise-botched-surgeries-misidentified-body-2026-02-09/.
Historical Consumer Price Index for All Urban Consumers (CPI-U): U.S. City Average, All Items, by Mon. (n.d.). Bureau of Labor Statistics. https://www.bls.gov/cpi/tables/supplemental-files/historical-cpi-u-202212.pdf.
Lamb, J. (2023, December 15). Radiological Computer Aided Triage and Notification Software. U.S. Food & Drug Administration. https://www.accessdata.fda.gov/cdrh_docs/pdf23/K233968.pdf.
Lång, K., Josefsson, V., Larsson, A.-M., Larsson, S., Högberg, C., Sartor, H., Hofvind, S., Andersson, I., & Rosso, A. (2023). Artificial Intelligence-Supported Screen Reading versus Standard Double Reading in the Mammography Screening with Artificial Intelligence Trial (MASAI): A Clinical Safety Analysis of a Randomised, Controlled, Non-Inferiority, Single-Blinded, Screening Accuracy Study. Lancet Oncology, 24(8), 936–944. National Library of Medicine. https://pubmed.ncbi.nlm.nih.gov/37541274/.
Levine, D. M., Ouchi, K., Blanchfield, B., Saenz, A., Burke, K., Paz, M., Diamond, K., Pu, C. T., & Schnipper, J. L. (2019). Hospital-Level Care at Home for Acutely Ill Adults: A Randomized Controlled Trial. Annals of Internal Medicine, 172(2), 77–85. National Library of Medicine. https://pubmed.ncbi.nlm.nih.gov/31842232/.
Microsoft Dragon Copilot. (n.d.). Microsoft. https://www.microsoft.com/en-us/health-solutions/clinical-workflow/dragon-copilot.
NCSBN Research Highlights Small Steps Toward Nursing Workforce Recovery; Burnout and Staffing Challenges Persist. (2025, April 17). National Council of State Boards of Nursing. https://www.ncsbn.org/news/ncsbn-research-highlights-small-steps-toward-nursing-workforce-recovery-burnout-and-staffing-challenges-persist.
NHE fact sheet. (2023, September 6). Www.cms.gov; U.S. Centers for Medicare & Medicaid Services. https://www.cms.gov/data-research/statistics-trends-and-reports/national-health-expenditure-data/nhe-fact-sheet.
Our Platform. (n.d.). Waystar. https://www.waystar.com/our-platform/denial-prevention-recovery/denial-appeal-management/.
The Complexities of Physician Supply and Demand: Projections From 2021 to 2036 Summary Report Prepared for the AAMC by GlobalData Plc. (2024). In Association of American Medical Colleges. https://www.aamc.org/media/75231/download.
The Financial Stability of America’s Hospitals and Health Systems Is at Risk as the Costs of Caring Continue to Rise. (2023, April). https://www.aha.org/guidesreports/2024-05-01-2023-costs-caring.
United States Health at a Glance. (2025). Organization for Economic Cooperation and Development. https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/11/health-at-a-glance-2025-country-notes_2f94481e/united-states_9139f48b/3517f35e-en.pdf.
Wong, A., Otles, E., Donnelly, J. P., Krumm, A., McCullough, J., DeTroyer-Cooley, O., Pestrue, J., Phillips, M., Konye, J., Penoza, C., Ghous, M., & Singh, K. (2021). External Validation of a Widely Implemented Proprietary Sepsis Prediction Model in Hospitalized Patients. JAMA Internal Medicine, 181(8). https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2781307.
Zweig, M., Kimmell, J., & Knowles, M. (2026, January 12). 2025 year-end digital health funding overview: A tale of two markets | Rock Health. Rock Health. https://rockhealth.com/insights/2025-year-end-digital-health-funding-overview-a-tale-of-two-markets/