Executives at Amazon, Ford and JPMorgan Chase have all predicted AI will eliminate a wide swath of white-collar jobs in the years ahead. Healthcare, for now, has mostly been an exception: Job openings in nearly every healthcare category tracked by Indeed remain above prepandemic levels, and the sector has largely avoided AI-linked layoffs hitting tech and marketing.
That does not mean nothing is changing. Federal labor projections and health system technology executives describe a narrower shift already underway inside health IT, revenue cycle and health information management departments. Below are five roles under the most pressure, and three doing similar-looking work that is growing instead.
Most at risk
1. Medical transcriptionists
The Bureau of Labor Statistics projects the occupation will shrink 4% by 2035, to about 40,200 jobs — one of the few healthcare-adjacent categories with an outright projected decline. BLS attributes the drop to advances in speech recognition and natural language processing that let physicians document patient encounters in real time.
2. Medical coders
Coding is moving in a similar direction, even though the broader medical records specialist category it sits inside is still predicted to grow 8% through 2035, driven mostly by an aging, sicker population that needs more medical services. BLS adds a caveat to that growth projection: increasing adoption of AI-powered coding tools “may affect the demand for these workers.”
Anthropic’s Claude usage data backs up both trends: medical records specialists and medical transcriptionists carry observed exposure scores of 0.667 and 0.637, respectively — surpassed only by data entry keyers and customer service representatives among all 750-plus occupations the AI company tracks.
3. Medical secretaries and administrative assistants
This role ranks as the single most AI-vulnerable healthcare occupation, according to a Washington Post analysis built on a GovAI and Brookings Institution vulnerability tool.
Sunil Dadlani, executive vice president and chief information, digital and cybersecurity officer of Morristown, N.J.-based Atlantic Health System, told Becker’s in 2025 that “tasks that are repetitive, rules-based, or data-heavy are the most vulnerable,” singling out documentation and scheduling among the functions already being “supported — or replaced” by intelligent systems.
4. Medical and health services managers
The administrators who typically oversee documentation-heavy departments land in the Post‘s “high vulnerability” tier as well — one of the more counterintuitive findings in that dataset, and one that doesn’t hold up against how the role is actually being used. Anthropic data puts the same occupation among the least AI-exposed in healthcare, with an observed exposure score of 0.066 — roughly a tenth of the score for medical records specialists or medical transcriptionists. The mismatch is a reminder that “vulnerable to AI” and “already being touched by AI” are two different questions, and this occupation scores high on the first without yet showing up much in the second.
The health system job cuts reported so far in 2026 bear that out only partially. Becker’s tracker of IT and coding layoffs shows Lewiston, Maine-based Central Maine Healthcare eliminating 38 IT positions tied to the retirement of legacy systems after an Epic go-live, and West Des Moines, Iowa-based UnityPoint Health cutting 207 IT and revenue cycle roles as it shifted the work to third-party vendors — consolidation and outsourcing, in other words, rather than AI directly. Portland, Maine-based MaineHealth cited advances in technology, among other reasons, when it eliminated 56 IT positions and consolidated three analytics teams into one, a net reduction of 27 roles, for a total of 83 eliminated positions.
5. Patient access and scheduling representatives
Kathy Azeez-Narain, chief digital and customer innovation officer of Newport Beach, Calif.-based Hoag, told Becker’s last year that her system is already automating “billing, call centers, and other repetitive workflows.” Customer service representatives carry an AI vulnerability score of 0.408 in Microsoft Research’s own analysis of Bing Copilot conversations — placing the role among the top 25 highest-scoring occupations of the more than 700 Microsoft tracked, alongside writers, editors and data scientists.
Holding up
Not every category built around codes and paperwork is shrinking. A smaller set of roles sitting right next to the ones above is positioned to grow instead.
1. Health information technologists and medical registrars
These analysts, who build and maintain the systems that coders and records specialists feed data into, are projected to grow 16% by 2035, to about 48,600 jobs — the fastest rate of any health IT-adjacent occupation BLS tracks, with no AI headwind noted in the agency’s outlook. The distinction BLS draws is telling: it is the data-entry and classification side of health information management under pressure, not the systems-design and analytics side.
2. AI governance roles
AI governance is emerging as a growth category in its own right. Eighty-four percent of health systems now have a dedicated AI governance committee, according to a CHIME Foundation and Censinet survey of 51 healthcare organizations released in December, with CIOs sitting on 63% of them and chief medical information officers on 45%. Only 8% of respondents described themselves as “very confident” in their ability to spot emerging AI risks — a gap health systems appear to be staffing for rather than folding into existing roles.
3. Cybersecurity and AI oversight roles
Several chief medical information officers named uncontrolled AI adoption among the biggest IT risks facing health systems in 2026, according to a recent Becker’s roundup. Usman Akhtar, MD, chief medical information officer of Arlington, Va.-based VHC Health, said growing reliance on EHRs, cloud platforms and integrated third-party applications means downtime now carries clinical consequences, not just financial ones. Amer Saati, MD, PhD, CMIO of Roseville, Calif.-based Adventist Health, separately warned against deploying AI without strong governance, validation and monitoring in place.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
