April 2, 2026 | Welby Health
The U.S. healthcare system is short roughly 84,000 physicians, 250,000 registered nurses, and 81,000 licensed practical nurses. More than 65% of hospitals have operated below full capacity at some point because they simply could not staff the beds. Two in five healthcare workers say their jobs feel unsustainable.
These numbers are not new. What is new is that every traditional lever for fixing them -- sign-on bonuses, travel nurse contracts, accelerated training pipelines -- has been pulled hard for years and the gap keeps widening. We are not going to recruit our way out of this. The pipeline of new clinicians cannot keep pace with the aging population, the rising acuity of chronic disease, and the administrative burden that drives experienced nurses and physicians out of the profession entirely.
So the question for every health system CEO, CFO, and CMO in 2026 is no longer "how do we find more people?" It is "how do we make the people we have dramatically more effective?"
That is an AI question. And the answer is more nuanced than most of the industry conversation acknowledges.
There is a version of the AI-in-healthcare narrative that treats technology as a substitute for clinical staff. That version is wrong, and organizations that chase it are going to get burned.
AI that supports clinicians is fundamentally different from AI that replaces them. The core of nursing, the bedside judgment, the patient relationship, the clinical intuition built over years of practice, cannot be automated. What can be automated is the mountain of administrative work that buries clinicians and drives them out of the profession.
Physicians spend an estimated two hours on administrative tasks for every one hour of direct patient care. Nurses lose up to 25% of their shifts to documentation. That is not a technology problem in the traditional sense. It is a design problem. We built healthcare workflows around documentation requirements instead of around patient care, and now we are surprised that the people doing the work are exhausted and leaving.
AI is the tool that can invert that equation, but only if it is deployed as a force multiplier rather than a headcount reduction strategy.
The most impactful AI deployments in healthcare right now are not the flashy diagnostic tools or the drug discovery platforms getting venture capital attention. They are the unglamorous workflow engines that eliminate repetitive clinical labor.
AI-powered ambient documentation tools are cutting charting time by up to 75%. Automated scheduling and patient matching systems are reducing staffing demand by 15% to 35%. Prior authorization automation is reclaiming hours per week for clinical staff who previously spent that time on hold with payer call centers.
BCG's research on AI in healthcare points to a useful framework: the 10-20-70 rule. Successful AI implementations dedicate 10% of effort to algorithms, 20% to technology and data infrastructure, and 70% to people and processes. That ratio matters because it reflects a truth the industry is still learning. The technology is not the hard part. Redesigning workflows so clinicians actually benefit from the technology is the hard part.
And this is where most health systems are falling short. AI adoption among U.S. healthcare providers has jumped from 34% in 2024 to a projected 68% in 2026. But adoption is not the same as integration. Nearly 90% of healthcare workers report using AI in some capacity, yet more than 60% of organizations still cite recruitment and retention as their biggest operational challenge. If the AI is deployed but the workforce crisis persists, the deployment is not solving the right problem.
Nowhere is the workforce multiplier more urgently needed than in chronic care management. More than 60% of American adults live with at least one chronic condition. These patients require ongoing monitoring, medication management, care coordination across multiple providers, and regular clinical touchpoints between office visits. That work is labor-intensive, repetitive in its structure, and critically important to outcomes.
It is also exactly the kind of work where AI-powered workflows paired with skilled clinicians produce outsized results.
At Welby, this is the model we built from the ground up. Our platform pairs licensed RN case managers with AI that automates vital sign monitoring, medication adherence tracking, and patient communication across chronic care management, remote patient monitoring, and transitional care programs. The AI handles the data ingestion, risk stratification, and workflow routing. The nurses handle the clinical judgment, patient relationships, and intervention decisions.
The results are not theoretical. Patients using our cellular-enabled blood pressure cuffs see a 20% decrease in blood pressure. Smart glucose monitors drive a 20% or greater reduction in blood glucose within four weeks. Heart failure patients on daily connected scales are 5.5x more likely to adhere to life-saving therapies.
Those outcomes do not happen because of the AI alone. They happen because the AI gives our clinical team leverage. Instead of spending their time pulling data from disparate systems, chasing down missing readings, or manually triaging alerts, our nurses spend their time doing what they were trained to do: delivering care.
Despite the evidence, fewer than one in four home-based and chronic care organizations have made AI-specific investments. Only about one-third are using technology-enabled training to address staffing gaps. Meanwhile, 60% of care-at-home leaders believe AI will have the greatest impact on their industry by 2030.
That gap between belief and action is where organizations are going to get separated. The ones investing now in AI-augmented care models are building the operational infrastructure that will let them scale as demand increases. The ones waiting for the technology to "mature" are going to find themselves competing for the same shrinking pool of clinicians with no force multiplier to show for it.
This is not a speculative argument. Value-based care arrangements are projected to cover 45% of Medicare beneficiaries by the end of 2026. Those arrangements reward outcomes, not volume. You cannot deliver better outcomes for a growing chronic disease population with a shrinking clinical workforce unless you fundamentally change how that workforce operates.
The healthcare workforce crisis is structural, not cyclical. It is not going to be solved by hiring, and it is not going to be solved by replacing clinicians with chatbots. It is going to be solved by organizations that figure out how to make every clinician two or three times more effective by pairing them with AI that handles the work that should never have been on their plate in the first place.
The organizations that treat AI as a force multiplier for their clinical teams will scale. The ones that treat it as either a silver bullet or a nice-to-have will not.
The math is clear. The question is whether your organization is acting on it.
TL;DR: The U.S. is short roughly 84,000 physicians and 250,000 RNs, and traditional hiring strategies cannot close that gap. AI adoption in healthcare has nearly doubled since 2024, but most organizations are deploying it without redesigning workflows around their clinical teams. The real opportunity is using AI as a force multiplier for clinicians, automating the administrative burden that drives burnout while letting nurses and physicians focus on patient care. In chronic care management, this model is already producing measurable outcomes. With value-based care covering 45% of Medicare beneficiaries by year-end, the organizations that pair AI with skilled clinicians will scale. The ones still trying to hire their way out of the crisis will not.
