By Greg Ikner, President, The AGA Group
AI and healthcare jobs have become part of a much larger conversation about the future of work. I have spent more than four decades in life sciences, business leadership and healthcare recruiting. During that time, I have watched technology change how people work more times than I can count.
Not simply because of what the technology can do, but because some of the people building the world’s most advanced AI systems are now asking whether its capabilities are advancing faster than our ability to safely manage them.
On September 12, Anthropic CEO Dario Amodei called for the industry to “pace the frontier.” His proposal was not to stop artificial intelligence development. Instead, he argued for slowing the advancement of frontier capabilities enough to give safety research, independent evaluation and oversight time to catch up. OpenAI CEO Sam Altman subsequently supported the idea of independent evaluators with deep access to AI companies.
That debate deserves attention well beyond Silicon Valley.
For healthcare, it raises a particularly important question about AI and healthcare jobs:
As artificial intelligence becomes more capable, what should healthcare continue to entrust to human beings?
AI Is Already Part of Healthcare
This is no longer a conversation about technology that might arrive someday.
According to the American Medical Association’s 2026 Physician Survey on Augmented Intelligence, 81% of physicians now report using AI professionally, more than twice the 38% reported in 2023. More than three-quarters say AI improves their ability to care for patients.
That is significant adoption in only a few years.
AI can summarize medical information, assist with documentation, analyze large amounts of data and reduce repetitive administrative work. It is also finding applications in medical research, drug development and clinical decision support.
Those benefits are real.
However, the AMA survey also reveals something equally important. Eighty-eight percent of physicians report at least some concern about AI-related skill loss. In addition, 85% want to be consulted or directly involved in decisions about AI adoption. Physicians also identified clear liability rules, privacy protections, and strong safety and efficacy validation as important conditions for broader adoption.
In other words, healthcare professionals are not rejecting AI.
They are asking reasonable questions about where it belongs.
Healthcare Raises the Stakes of Artificial Intelligence
An incorrect AI-generated marketing recommendation can be inconvenient.
An incorrect recommendation involving a patient’s health can have considerably greater consequences.
That distinction helps explain why the U.S. Food and Drug Administration is examining generative AI-enabled medical devices differently from conventional software.
In August 2026, the FDA released a discussion paper seeking input on risk assessment, premarket evaluation and postmarket monitoring of generative AI-enabled medical devices. Importantly, the FDA makes clear that the document is for discussion only. It is not draft guidance, final guidance or a proposed policy change.
The distinction matters because generative AI can accept open-ended information and produce variable responses. That creates different questions about reliability, monitoring and accountability.
The World Health Organization has raised similar concerns.
WHO guidance identifies potential benefits for diagnosis, clinical care, administrative work, medical education and scientific research. At the same time, it warns about inaccurate information, bias, privacy, cybersecurity and what it calls “automation bias,” where people may overlook errors because they trust the technology.
More recently, WHO’s 2026 work on AI and health policy made the principle especially clear: AI should augment rather than replace human judgment, with human oversight remaining part of responsible implementation.
That strikes me as an important distinction.
The issue is not whether AI belongs in healthcare. It already does.
The question is how much responsibility we are prepared to give it.
AI and Healthcare Jobs Are More Complicated Than “Replacement”
Much of the public conversation about artificial intelligence eventually arrives at the same question:
Will AI take my job?
The evidence does not support a simple answer.
Anthropic’s own March 2026 labor-market research found no systematic increase in unemployment among workers in occupations highly exposed to AI since late 2022. However, researchers found suggestive evidence that hiring of younger workers has slowed in more exposed occupations. Anthropic itself describes the labor-market transition as uncertain.
What appears increasingly likely is that AI will affect tasks within jobs before it eliminates entire occupations.
Healthcare offers a good example.
AI can already help draft notes, summarize information and perform portions of administrative work. That could reduce the amount of time people spend on repetitive tasks.
Some administrative positions built primarily around predictable, rules-based work may face greater pressure. Other positions will change as employees move from producing information themselves to reviewing, correcting and acting on AI-generated information.
But that is very different from saying healthcare will need fewer people.
Healthcare Still Needs People
The latest Bureau of Labor Statistics projections provide an important counterweight to predictions of widespread healthcare job loss.
From 2025 through 2035, private healthcare and social assistance is projected to add more than 2.2 million jobs, more than any other major industry sector. Healthcare support occupations are projected to grow 13.3%, while healthcare practitioner and technical occupations are projected to grow 8%. Together, those two healthcare groups are expected to account for almost one-third of all new jobs created through 2035.
Nurse practitioners are projected to grow 41%. Medical and health services managers are projected to grow 24.2%. Registered nurses are projected to have about 180,800 openings annually.
Why?
Because technology does not eliminate the underlying reasons people need healthcare.
The population continues to age. Chronic disease remains prevalent. People still require diagnosis, treatment, rehabilitation, preventive care and support.
AI may change how that care is delivered.
It does not make the patient disappear.
The Human Skills That May Become More Valuable
There is another possibility that receives less attention.
As artificial intelligence becomes better at processing information, some distinctly human abilities may become more valuable, not less.
Healthcare requires judgment when the answer is unclear.
It requires accountability when a decision has consequences.
It requires someone to recognize when the data does not match what they are seeing in the patient sitting in front of them.
Healthcare also requires communication.
A machine may be able to analyze a diagnosis. Someone still has to explain that diagnosis to a frightened patient or family.
Technology can identify patterns. Someone still has to decide what those patterns mean in the context of an individual life.
And much of healthcare remains physical.
Someone still examines the patient, draws the blood, provides the therapy, cleans the teeth, administers the medication, assists the physician, comforts the family and notices when something simply does not look right.
Those are not insignificant details.
They are healthcare.
What AI Should Do and What People Should Do
Perhaps healthcare is asking the wrong question when it asks whether AI will replace people.
A better question may be:
What should technology do, what should people do, and where should they work together?
AI can process extraordinary amounts of information. It can recognize patterns humans might miss. It can automate repetitive work and potentially give clinicians and administrators more time for higher-value responsibilities.
Those capabilities should not be dismissed simply because the technology also presents risks.
At the same time, efficiency cannot be the only measure of progress in healthcare.
Accuracy matters.
Accountability matters.
Trust matters.
Judgment matters.
And people matter.
The Future of AI and Healthcare Jobs
I do not know exactly what healthcare will look like ten years from now. I am skeptical of anyone who says they do.
Artificial intelligence is moving too quickly, and even the people building these systems continue to revise their expectations about what comes next.
But there is something we can see more clearly.
The future of healthcare is unlikely to be a choice between technology and people.
It will be about understanding what each does best.
AI will become more capable. Some healthcare jobs will change. Some tasks will disappear. New responsibilities will emerge, and some positions may eventually disappear as well.
Yet healthcare will still require someone to take responsibility.
Someone will still have to listen.
Someone will still have to explain.
Someone will still have to make the difficult judgment when the information does not provide a perfect answer.
And someone will still have to care for the person, not simply process the patient.
Perhaps that is the larger lesson in the current debate about artificial intelligence.
The more capable our technology becomes, the more important it may become to understand what should remain human.
In healthcare, that distinction matters more than almost anywhere else.
