On January 1, 2015, a billing code changed American primary care.
CPT 99490 — chronic care management services — authorized Medicare reimbursement for 20 minutes of non-face-to-face clinical staff time spent coordinating care for patients with two or more chronic conditions. Before that date, care coordination was work that everyone agreed was important, that many clinicians performed informally, and that no one got paid for. After that date, it was a billable service.
Within two years, health systems across the country had built dedicated care coordination departments. They hired RNs, LPNs, and social workers into a new role — care coordinator — that hadn't existed as a formal position in most organizations before the code was published. The reimbursement code didn't just pay for existing work. It created a workforce.
Eleven years later, that same dynamic is beginning with AI — and almost nobody in the career-advice ecosystem is paying attention to it.
The technologists are focused on what AI can do. The futurists are focused on what AI will do. The people who will actually determine what AI does to healthcare careers are doing something much less glamorous: they're writing reimbursement codes at CMS and CPT codes at the AMA. In American healthcare, money is the operating system. No clinical AI tool gets widely deployed until someone figures out how to bill for it. No role gets created or eliminated until the economics support it. And the economics of healthcare run through the billing code.
The clinician who wants to understand how AI will reshape their career should spend less time reading about GPT-5 and more time reading the Federal Register.
How Billing Codes Create Workforce
This claim — that billing codes create jobs — sounds reductive. It isn't. It's the most reliable predictor of healthcare workforce change in the United States, and it has been for decades.
The mechanism is straightforward: when CMS creates a reimbursement pathway for a service, health systems build the operational infrastructure to deliver that service. The infrastructure requires people. The people become a workforce. The workforce eventually gets credentials, training programs, and professional associations. The whole chain starts with the code.
99490 created care coordination. Before 2015, care coordination was an aspiration. After 2015, it was a revenue-generating service. Health systems staffed it accordingly. Today, care coordination is a recognized professional pathway with dedicated roles, a CCM credential from CCMC, and a labor market that the BLS tracks.
Remote patient monitoring codes created RPM programs. CPT 99453, 99454, 99457, and 99458 — collectively, the RPM code family — authorized reimbursement for device setup, data transmission, and treatment management services for patients monitored remotely. The 2026 Medicare Physician Fee Schedule expanded this further with CPT 99445 (shorter-duration monitoring) and new treatment management codes (99470, 98979) for 10–19 minutes of monthly service. Each new code lowers the threshold for programs to be economically viable — which means more programs, which means more staff. CMS has now extended telehealth flexibilities through 2029. That's not a temporary pandemic policy. That's a four-year signal to build permanent infrastructure.
Meaningful Use incentives created the HIT workforce. The HITECH Act didn't create a billing code — it created something more powerful: $27 billion in direct financial incentives for EHR adoption. The workforce response was immediate. Between 2009 and 2012, healthcare IT job postings increased 199%. ONC's Workforce Development Program trained 19,733 graduates through 82 community colleges for roles that didn't exist four years earlier. More than 50,000 new positions were created. The money came first. The jobs followed.
The pattern is not subtle. It's not speculative. It's the documented history of how healthcare workforce changes actually happen in the United States.
The AI Billing Codes Are Here
For years, the knock on AI in healthcare was that there was no way to get paid for it. That objection is now obsolete.
The CPT 2026 code set, effective January 1, 2026, includes 288 new codes — and for the first time, multiple codes specifically cover AI-augmented clinical services. The AMA has built a framework for coding AI-enabled services under the term "augmentative" — defined as AI that "analyzes or quantifies data in a clinically meaningful way, but still requires interpretation by a physician or other qualified health care professional."
The initial AI-augmented codes are concentrated in imaging and diagnostics:
- Coronary atherosclerotic plaque assessment — AI-derived analysis of coronary CT angiography data to assess disease severity
- Perivascular fat analysis for cardiac risk — AI-derived assessment of cardiac risk from perivascular fat analysis with or without concurrent CT
- AI-augmented radiology interpretation — Category III codes for AI-assisted detection and interpretation across imaging modalities
These are not future proposals. They are active, billable codes in the 2026 fee schedule. They establish the principle that AI-augmented clinical work is a reimbursable service — and that the physician who interprets the AI's output is the one who bills for it.
This is the beginning, not the end. The AMA has explicitly stated that CPT codes now "offer the language to report AI-enabled health services." The infrastructure for billing AI-assisted work across specialties is being built. As it expands, the financial incentive to deploy AI tools — and to staff the roles required to use them properly — will grow.
Every one of those codes creates a downstream workforce question: who performs the service, who supervises the AI, who validates the output, and who gets paid? The answers to those questions will determine which clinical roles expand and which contract over the next decade.
The Prior Authorization Earthquake
If AI billing codes are the slow-building wave, prior authorization reform is the earthquake.
The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) requires Medicare Advantage organizations, state Medicaid and CHIP programs, and managed care plans to respond to prior authorization requests within 72 hours for urgent cases and seven calendar days for standard cases — effective January 1, 2026. API requirements for automated prior authorization processing take effect January 1, 2027.
To meet these timelines at scale, payers are deploying AI-automated prior authorization systems. Some are already auto-approving up to 90% of requests. The industry estimates $13.3 billion in annual administrative savings from full automation of prior authorization, according to CAQH Index data. One revenue cycle organization reported 15,000 saved employee hours per month from automation.
This is not a technology story. It is a workforce story.
There are currently hundreds of thousands of people in the United States whose primary job function is processing prior authorization requests — on both the provider and payer sides. Nurses, medical assistants, and administrative staff at physician practices spend hours daily on phone calls, faxes, and portal submissions. Utilization review nurses and medical directors at health plans spend their days reviewing requests and applying clinical criteria.
When AI automates 90% of that workflow, those people don't disappear. They get redeployed. And where they get redeployed is a career-strategy question that almost nobody is asking.
On the provider side, the clinical staff currently consumed by prior authorization will be reallocated. Some will move to care coordination — a function that continues to expand as RPM and CCM codes broaden. Some will move to AI oversight roles — reviewing the cases that the automated system flags as exceptions. Some will be absorbed into population health management, where AI-generated patient risk stratification creates demand for human follow-up.
On the payer side, utilization review nurses face a more direct disruption. The clinical review function that defines their role is being automated. The pivot is toward exception handling, AI audit, appeals management, and the complex medical necessity determinations that automated systems can't resolve.
The prior authorization rule didn't mention AI by name. It doesn't need to. The timelines it mandates are impossible to meet without AI, and the workforce consequences will be determined by how organizations redeploy the staff that AI displaces from the prior auth workflow. That redeployment is a career opportunity for clinicians who see it coming and position themselves accordingly.
The Value-Based Care Accelerant
The third billing-code dynamic reshaping healthcare careers is the continued expansion of value-based payment models — and the role AI plays in making them profitable.
For Performance Year 2026, 511 ACOs are participating in the Medicare Shared Savings Program — up from 476 in 2025 — covering more than 700,000 healthcare providers serving 12.6 million Medicare beneficiaries, a 12.3% increase from the prior year. CMS has set a goal of having all Traditional Medicare beneficiaries in a quality-focused care relationship by 2030.
Under value-based contracts, health systems generate shared savings revenue by improving quality metrics and reducing total cost of care. AI tools are increasingly the mechanism for achieving both: predictive models identify high-risk patients before they deteriorate, AI-generated care gap reports flag overdue screenings, population health dashboards surface the patterns that enable proactive intervention.
Here's the workforce implication: every AI tool that improves a quality metric under a value-based contract generates revenue. That revenue funds headcount. And the headcount it funds is concentrated in the roles that make AI tools operational — care coordinators who act on risk stratification outputs, data analysts who monitor population health dashboards, informaticists who configure and validate the AI models, and quality improvement staff who track metric performance.
The quality reporting requirements are expanding in lockstep. ACOs will report on eight quality measures in 2026, up from six in 2025, increasing to nine in 2027 and eleven in 2028. Each additional measure creates additional analytical, coordination, and reporting work. Each additional ACO participating in shared savings creates additional demand for the professionals who make value-based care operationally viable.
The roles that value-based care rewards are not the roles that fee-for-service rewarded. Fee-for-service rewarded volume — more visits, more procedures, more billing events. Value-based care rewards outcomes — fewer readmissions, better chronic disease control, higher screening rates. The workforce that delivers outcomes is different from the workforce that delivers volume. It includes care coordinators, population health analysts, quality improvement specialists, and clinical informaticists. The billing code structure is funding their expansion.
A Framework for Reading Career Signals in Billing Policy
For the clinician, student, or career-changer who wants to use this analysis practically, here's a framework for translating billing policy into career strategy:
1. When CMS creates a new reimbursement code, ask: "Who does this work create?"
Every new CPT code funds a workflow. Every funded workflow requires staff. When you see a new code — whether it's an AI-augmented radiology interpretation, a shorter-duration RPM monitoring code, or an expanded CCM service — trace the operational chain: who performs the service, who supervises it, who documents it, who bills for it? The roles at the end of that chain are the ones that will grow.
2. When CMS mandates a timeline, ask: "What technology makes this possible, and who operates it?"
The prior authorization rule mandates 72-hour response times. The technology that makes it possible is AI-automated review. The people who operate it are the exception handlers, the AI auditors, and the clinicians who manage the cases too complex for automation. A mandate creates a technology requirement. A technology requirement creates an operational need. An operational need creates a role.
3. When CMS expands a value-based program, ask: "What roles make the quality metrics move?"
Shared savings revenue depends on quality performance. Quality performance depends on care coordination, data analysis, population health management, and clinical decision support. When ACO participation grows — as it is, by 12.3% year-over-year — the demand for these roles grows with it. The growth won't show up in a BLS projection for five years. It's visible in the CMS participation data today.
4. When the AMA adds codes for AI-augmented services, ask: "What clinical specialty does this make more valuable?"
The first AI-augmented billing codes are in radiology and cardiology. They authorize payment for the physician who interprets AI-generated output. They don't pay the AI. They pay the human who validates what the AI produced and takes professional responsibility for the result. The specialties that get AI-augmented codes first are the ones where AI-human collaboration becomes a billable, revenue-generating activity — and where the physician's interpretive role is formally codified as worth paying for.
5. Follow the proposed rules, not just the final rules.
CMS publishes proposed rules 6–12 months before they take effect. The Federal Register is publicly available. The proposed rules are the earliest signal of where the money is going — and where the money goes, the workforce follows. A proposed rule for AI-assisted diagnostic billing is a career signal. A proposed expansion of RPM codes is a hiring forecast. Reading proposed rules is due diligence for your career, the same way reading earnings calls is due diligence for an investment portfolio.
Who Wins, Who Pivots, Who Acts
The billing code landscape that's taking shape in 2025–2026 has clear implications for specific career pathways:
Medical coders and HIM specialists face the most direct pressure. AI coding tools are in production and their accuracy is approaching human levels for routine cases. But the new AI-augmented billing codes create compliance complexity that requires human expertise — someone has to ensure that the AI-augmented service was documented correctly, that the physician's interpretive role is substantiated, and that the code assignment reflects what actually happened. The coder who becomes an AI coding auditor — who understands both the rules and the technology — has a durable role.
Care coordinators are on the growth side of every policy signal. CCM codes are expanding. RPM codes are expanding. ACO participation is expanding. Value-based care is expanding. Every one of these trends increases demand for the professionals who coordinate, follow up, and manage the gaps between visits. The billing code trajectory for care coordination points unambiguously toward growth.
Clinical informaticists are positioned at the intersection of every AI billing development. The new AI-augmented codes require that AI tools be properly deployed, validated, and documented within clinical workflows. That's informatics work. AI governance, vendor evaluation, and model monitoring — all of which the billing codes incentivize — are informatics functions. AMIA estimates demand growing at 2x the training pipeline. The billing codes are about to steepen that curve.
Health data analysts are the operational backbone of value-based care. Every quality measure, every population health dashboard, every risk stratification model requires an analyst who can build it, monitor it, and explain it to clinical and executive leadership. The expansion from six to eleven ACO quality measures over three years is, in practical terms, a hiring signal for analysts.
Nurse practitioners stand to benefit from the AI-augmented primary care codes — if they come, which the current trajectory suggests they will. AI tools that extend an NP's capacity in primary care are worth more when there's a billing code for the AI-assisted service. Full practice authority in 26+ states, combined with AI augmentation, combined with reimbursement codes, creates a compounding growth dynamic for the NP role.
The Federal Register as Career Tool
The punchline of this analysis is uncomfortable for people who prefer their career advice to come from LinkedIn influencers and conference keynotes: the single most useful document for understanding how AI will reshape healthcare careers is the Federal Register.
It's not intuitive. It's not glamorous. It's a government publication written in regulatory prose, published on a rolling basis, available free at federalregister.gov. But it contains the proposed and final rules that determine where $1.5 trillion in annual federal healthcare spending goes — and that spending is what creates, sustains, and eliminates the jobs that clinicians hold.
The clinician who reads the Federal Register — or at least reads the summaries that AAFP, AMA, AMIA, and AHIMA publish within days of each major rule — has a structural advantage over the clinician who reads AI headlines. The headlines tell you what the technology can do. The billing codes tell you what the technology will do — because they tell you what someone will pay for.
Follow the money. In American healthcare, the money follows the billing code.
LumenHealth helps healthcare organizations build AI governance frameworks that match their risk, scale, and mission. Take the assessment to see where you stand.
Assess your organization's AI governance readiness
37 questions across five domains. Free facilitated debrief with your leadership team.