Aug 2026· NEJM catalyst innovations in care delivery· Vol 7 9, pp.
CAT260037
· 0 citations· 37 references
Medicine
TL;DR
The analysis considered the savings potential from full national implementation of three specific AI technologies - machine learning (ML), natural language processing (NLP), and generative AI (genAI) - across administrative and medical expense categories for payers and providers (but excluded onetime implementation costs).
Abstract
Artificial intelligence (AI) offers the potential to improve productivity and reduce waste across the U.S. health care system. Quantifying achievable value from AI technologies can inform organizational strategies and national spending projections. This study aimed to estimate the annual run-rate net value achievable within 5 years through full adoption of AI use cases across major health care stakeholders and domains without compromising quality or access. This study applied observed implementation benchmarks and national expenditure data to 12 AI-enabled domains spanning five stakeholder groups - private payers, public payers, hospitals, physician groups, and other sites of care - collected during the period of October 2023 and March 2024. The financial impact was estimated using 2024 data (the latest year for which full data are available) of U.S. health care expenditures, using ranges from published literature and observed implementation evidence. No human participants were involved. The analysis considered the savings potential from full national implementation of three specific AI technologies - machine learning (ML), natural language processing (NLP), and generative AI (genAI) - across administrative and medical expense categories for payers and providers (but excluded onetime implementation costs). Annual net value (2024 U.S. dollars) and percentage reductions in total, administrative, and medical expenses by stakeholder group and AI technology were assessed. Assuming full adoption - i.e., a health care environment in which all stakeholders are all-in on AI adoption for all domains across administrative and medical expenses for all designated AI technologies, in this case, ML, NLP, and genAI - AI could generate US$438.9-US$810.7 billion in annual net value (5.7%-10.6% of the US$7.7 trillion in 2024 total health care expenses). GenAI-led use cases account for 54.4%-55.9% of the total AI opportunity. Administrative expenses could decline by US$120.1-US$252.0 billion (9.4%-19.8%) and medical expenses by US$318.8-US$558.6 billion (5.0%-8.7%). By stakeholder group, estimated annual values are as follows: private payers, US$205.1-US$357.0 billion (7.0%-12.1%); hospitals, US$130.6-US$219.5 billion (8.6%-14.4%); physician groups, US$29.0-US$89.9 billion (2.8%-8.6%); public payers, US$62.4-US$107.4 billion (5.1%-8.8%); and other sites of care, US$11.9-US$36.9 billion (1.3%-4.0%). Financial impact is concentrated in health care management, provider relationship management, and claims management for payers and in clinical operations and quality and safety for providers. Labor productivity and administrative automation account for the largest share of impact. Full implementation of these AI technologies could reduce U.S. health care spending by up to US$810.7 billion within 5 years without compromising quality and access. Realizing this full potential would require properly aligned incentive models (e.g., between physicians and the hospital), as well as coordinated organizational change, workflow redesign, and infrastructure investment, especially in clinical domains. Responsible scaled adoption supported by policy and industry efforts could help bend the U.S. health care cost curve.
This article critically examines the potential impacts of artificial intelligence (AI) on prices, total spending, and the rate of spending growth in the areas of prescription drug innovation and expanded patient access to care - including developments in remote patient monitoring, chronic care management, direct-to-consumer health care, clinical decision support, and nonclinical administrative labor. Based on observed industry practices, economics, and public policy, the article presents a thought exercise on the most likely effects for patients, providers, payers, and the broader health care system. The authors argue that under the still-dominant fee-for-service payment model, as well as the highly consolidated hospital and insurance markets, AI is more likely to increase total costs and spending growth in the short to medium term rather than slow them, even as it delivers substantial access and clinical quality improvements for patients. The cost-bending potential of AI varies distinctly by fee-for-service versus value-based payment. Regulators need to introduce policy and reimbursement levers for AI to slow cost growth. The authors conclude that, without significant changes in the health care system's financial incentives and market structures, AI will not slow cost growth.
B. Kocher, Brian Zhao, Erin Duffy· NEJM catalyst innovations in...· 0 citations
INTRODUCTION
Health economic models (HEMs) provide a solid foundation for reimbursement policy decisions that shape patient access to new treatments and the allocation of scarce healthcare resources. Model development is labor-intensive and time-consuming, often requiring months of expert work. Recent advances in large language models (LLMs) prompted interest in whether artificial intelligence can support or partially automate this process, but the evidence base remains scattered and has not been mapped against the modeling workflow.
AREAS COVERED
This review examines current applications of LLMs to health economic modeling. Five proof-of-concept studies are included and mapped to an eight-stage workflow adapted from the ISPOR-SMDM Modeling Good Research Practices framework and discussed in terms of reproducibility, validation, adaptability, and technology readiness. Published work addressed model parameterization, model implementation, reporting and quality assessment, and local adaptation, while research question design, model conceptualization, uncertainty analysis, and model validation remained unaddressed.
EXPERT COMMENTARY
The evidence supports cautious optimism. Near-term gains are augmenting human modelers on decomposed, verifiable sub-tasks rather than pursuing autonomous end-to-end modeling, which remains distant given current reliability levels and the iterative, collaborative nature of model development.
Attila Imre, B. Németh, Á. Jóźwiak et al.· Expert review of pharmacoeco...· 0 citations
Artificial intelligence (AI) is increasingly being adopted across healthcare systems to improve diagnostic accuracy, streamline administrative processes, strengthen clinical decision-making, and support preventive care. However, its national economic implications remain insufficiently understood, particularly regarding healthcare cost savings, workforce productivity, and long-term population-health outcomes. This study develops an evidence-informed national economic modelling framework to estimate the potential impact of AI adoption across key healthcare functions, including diagnostic imaging, automated screening, clinical decision support, predictive risk management, telehealth, and administrative automation. The framework compares current healthcare delivery with low-, moderate-, and high-adoption AI scenarios over a ten-year period. Economic outcomes include direct medical cost savings, reduced avoidable hospitalizations, improved diagnostic efficiency, clinician and administrative time savings, increased service capacity, and long-term health benefits measured through avoided complications and quality-adjusted life years. The analysis also incorporates implementation, maintenance, workforce-training, data-infrastructure, and governance costs to estimate net economic value. The study argues that AI can generate substantial national value when deployed in high-volume, high-cost, and prevention-oriented services. Nevertheless, financial benefits depend on interoperable health-data systems, clinical integration, algorithmic accuracy, human oversight, equity safeguards, and continuous performance monitoring. The proposed framework offers policymakers a practical basis for prioritizing responsible AI investments that improve both healthcare efficiency and long-term patient outcomes.
T. Nguyen· International Journal of Art...· 0 citations
Importance
Health care policies often fail to achieve their goals due to implementation challenges attributable to workforce constraints, fragmented health information systems, and administrative complexity. This Special Communication proposes a framework for how artificial intelligence (AI) tools could support effective health care policy implementation, using the implementation of Medicaid work requirements under the Budget Reconciliation Act of 2025 as an example.
Observations
Opportunities for AI-augmented health care policy implementation include executing key policy processes, such as generating eligibility screening tools, reviewing documentation, and linking and analyzing data for indicators of policy compliance; identifying individuals at risk of adverse consequences from implementation failure who should receive proactive support; enhancing policy communication to diverse audiences; facilitating implementation monitoring; and learning from and adapting implementation across all of these domains. To realize the potential of AI augmentation, the field needs to overcome challenges related to data availability, as well as the limitations of AI tools themselves, such as hallucinated false information.
Conclusions and Relevance
AI-augmented health care policy implementation has the potential to meaningfully limit unintended consequences from implementation of Medicaid work requirements. Rigorous evaluation of state-led innovations in AI-augmented implementation of Medicaid work requirements is key to advancing effective approaches and mitigating the potential for harm. The federal government should support state efforts with AI expertise, data infrastructure, and partnerships with preferred vendors who demonstrate that their AI-augmented digital assistants and other AI tools effectively facilitate Medicaid enrollment among eligible individuals.
E. McGinty, Yongkang Zhang, Fei Wang et al.· JAMA Health Forum· 0 citations
Artificial intelligence is rapidly changing the way health systems deliver care, generate knowledge, and support decision making. Across the world, AI is being used to strengthen disease surveillance, improve diagnostic accuracy, accelerate drug discovery, and expand access to healthcare through digital platforms. These developments present important opportunities for Africa, where persistent shortages of healthcare workers, growing disease burdens, and unequal access to specialist services continue to challenge health systems.
At the same time, the benefits of AI will not be realised automatically. Without deliberate investment in digital infrastructure, local research capacity, data governance, ethical regulation, and workforce development, there is a real risk that Africa will remain primarily a consumer of technologies designed elsewhere. Such an outcome could deepen existing inequalities and limit the continent's ability to shape technologies that reflect its own health priorities and cultural contexts.
This commentary argues that the future of AI in African healthcare should be guided by local leadership, interdisciplinary collaboration, and equitable partnerships. It calls for governments, universities, researchers, healthcare institutions, and the private sector to work together to build an innovation ecosystem in which artificial intelligence strengthens health systems while advancing scientific independence. Africa's role in the AI era should extend beyond technology adoption to active leadership in developing solutions that contribute to both regional and global health
E. Elliason· Interdisciplinary Journal of...· 0 citations
Artificial intelligence (AI) is increasingly embedded in healthcare delivery, yet its evaluation remains dominated by technical performance metrics that inadequately capture real-world system value. Clinical accuracy is necessary but insufficient to justify adoption in resource-constrained health systems facing workforce shortages and rising demand. AI technologies compete for limited financial, technical and cognitive resources, but there is no standardised framework to assess their effects on productivity, workflow, downstream utilisation or equity. This article argues that treating clinical validity as a proxy for value risks misallocating scarce resources and undermining trust in digital transformation. We propose an AI Productivity Index to complement existing safety, efficacy and economic assessments by evaluating operational impact, implementation burden, opportunity costs and post-deployment consequences. Embedding productivity measurement into procurement and governance processes could help align AI innovation with fiscal accountability, equitable access and sustainable healthcare delivery.
Y. Al-Ajlouni, Basile Njei· Clinical medicine (London)· 0 citations