US healthcare leaders are deploying artificial intelligence faster than any other sector in the country. The workforce charged with using those tools is not keeping pace, and the gap is starting to show up in clinical workflows, patient care, and operational margins.
The People Readiness Report from Kyndryl, a New York-based enterprise technology services firm, found that nearly all US healthcare organizations are now deploying AI across clinical, operational, and administrative functions, a rate that outpaces the global average. The findings were released Thursday, May 21, through a guest article by Anupama Shashank, Managing Director and Senior Vice President for Healthcare and Life Sciences at Kyndryl, published in Healthcare IT Today.
Shashank’s core argument is one that healthcare executives across the country are now being forced to confront: investment in AI technology alone does not produce returns. Without parallel investment in the people expected to work alongside that technology, even sophisticated deployments stall before they scale.
The Gap Between Ambition and Adoption
The Kyndryl analysis describes a pattern that has become familiar inside major US hospital systems. AI tools are being deployed broadly, but adoption is uneven. Clinicians experience alert fatigue from systems that surface too many notifications without sufficient prioritization. Decision-making slows as physicians second-guess algorithmic outputs they were not trained to interpret. Operational friction increases in environments that are already among the most cognitively demanding in any industry.
“Without proper enablement, AI deployments risk undermining their own goals of reducing administrative burden and improving patient outcomes,” Shashank wrote in the report.
The numbers from Kyndryl’s earlier Healthcare Readiness Report, published in March 2026, set the structural backdrop. That earlier study found that 76% of healthcare organizations have more AI pilots than they can scale, and only 30% feel prepared to adapt to evolving policy and regulation, even though 55% report concerns about keeping pace with the regulatory environment. Roughly 31% identified regulatory and compliance issues as a key barrier to enterprise-wide deployment.
The People Readiness Report extends that diagnosis from the governance layer down to the workforce layer — and the picture that emerges is of a sector that has invested in technology without sufficiently investing in the change management required to make that technology work.
A State-by-State Regulatory Puzzle
The challenge sits on top of a fragmenting US regulatory landscape that gives compliance officers at major health systems no single playbook to follow. Utah’s AI disclosure laws have been in force since May 2025, with penalties of $2,500 per violation in regulated sectors including healthcare. Texas implemented sweeping plain-language disclosure requirements on January 1, 2026, applying to any AI-influenced “high-risk” scenario including healthcare and hiring. Colorado’s AI Act enforcement begins June 30, 2026, with annual impact assessments and anti-bias controls required for high-risk decisions. California’s rules require generative AI developers to disclose training data sources, apply watermarking, and issue disclaimers in health communications.
For executives at multistate hospital systems, that patchwork forces a defensive posture: build compliance infrastructure to satisfy the strictest applicable state — often Colorado or California — and treat every clinical AI deployment as a regulated workflow rather than a pilot. The Kyndryl data suggests that pressure is being absorbed by leadership without being fully translated into frontline operational protocols.
How Major Hospital Systems Are Responding
The leadership challenge cuts across the largest US providers. HCA Healthcare, the country’s biggest investor-owned hospital chain, has expanded its partnership with Google Cloud and centralized AI governance under a chief data and analytics officer. CommonSpirit Health, formed from the merger of Catholic Health Initiatives and Dignity Health, has invested in clinician-facing ambient documentation tools designed to reduce note-taking burden — a category where workforce adoption tends to be strongest because the value to the user is immediate. Kaiser Permanente has invested in clinical decision support and is building internal AI literacy programs across its medical groups. Cleveland Clinic has structured its AI deployment around a model of clinician co-design, embedding physicians and nurses in tool development rather than rolling out finished products into busy units.
These approaches share a common premise that the Kyndryl report formalizes: the path from AI investment to clinical outcome runs through workforce enablement, not around it. Hospital systems that treat training, change management, and feedback loops as a parallel workstream alongside technology deployment tend to scale faster. Systems that treat workforce readiness as a downstream consequence of procurement decisions tend to stall.
The Governance Layer Catches Up Last
Kyndryl’s broader 2026 healthcare strategy points to where the next layer of investment is likely to land. The company has been promoting a “policy as code” capability, described by Christine Landry, Global Vice President for Healthcare at Kyndryl Consult, as a tool for translating regulatory, security, and organizational policies directly into digital and AI systems. The framing matters: it positions compliance not as a manual review process layered on top of AI deployments, but as a structural feature built into them from the start.
Kyndryl has also announced collaborations with the Servei de Salut de les Illes Balear in Spain on a clinical-genomic AI platform, and with the University of Liverpool’s Civic Health Innovation Labs to apply its Agentic AI Framework to patient engagement workloads. Both initiatives are designed to push AI past the pilot stage and into scaled operational use.
For US healthcare leaders, the takeaway from the People Readiness Report is structural rather than tactical. The technology is ready. The regulation is multiplying. The capital is flowing. The variable that will separate the systems that scale AI safely from those that don’t is the workforce itself — and the leadership decisions that determine whether frontline staff are equipped to use the tools that have been bought for them.
