Across hospitals and clinics, a quiet paradox is taking shape: the healers most eager to welcome artificial intelligence into medicine are, by their own admission, among the least prepared to wield it wisely. A survey of 148 healthcare professionals from Germany and 17 other nations finds that enthusiasm for AI runs 56 percentage points ahead of actual knowledge — a gap that mirrors a broader human tendency to embrace transformation before understanding it. The moment calls not for dampened hope, but for institutions to meet the hunger their professionals are already feeling.
Healthcare Professionals Embrace AI With Enthusiasm But Lack Knowledge, Study Finds
Only 20 percent felt well-informed about healthcare AI, despite 76 percent expressing excitement.
Mark
Why does it matter that doctors are excited about AI if they don't understand it?
Mimi
Because excitement without knowledge creates a specific kind of danger. A clinician who believes in AI's potential but doesn't grasp its limitations might over-rely on an algorithm's output, even when their own clinical judgment contradicts it. That's called automation bias. It happens quietly.
Mark
But surely doctors are trained to be skeptical, to question things?
Mimi
They are. But that skepticism erodes when you're exhausted, when you trust the institution that deployed the tool, when you don't know enough to ask the right questions. The study found that only 20 percent feel well-informed about healthcare AI. You can't be appropriately skeptical about something you don't understand.
Mark
The survey shows institutions aren't funding training. Why would they do that if they know there's a gap?
Mimi
Many institutions aren't even having the conversation yet. A third of respondents said AI isn't part of their organization's planning at all. The ones that are moving forward often lack the resources or expertise to build training programs. And there's a chicken-and-egg problem: leadership doesn't understand AI well enough to fund training for it.
Mark
What surprised you most in the data?
Mimi
That the gap was so consistent. I expected younger doctors to be more confident, or that German healthcare—which is well-resourced—would have solved this. But the knowledge-enthusiasm gap appeared everywhere, in every age group, in every country represented. It's not a fringe problem. It's structural.
Mark
If professionals are willing to invest their own time in learning, why isn't that enough?
Mimi
Because individual effort can't substitute for institutional infrastructure. You can't build safe AI governance through personal initiative. You need standards, oversight, liability frameworks, integration with existing systems. A doctor learning ChatGPT on their own time is not the same as a health system deploying validated AI tools with proper safeguards.
Mark
What would actually fix this?
Mimi
Structured programs built into medical education and continuing training. Hands-on experience with real tools in controlled settings. Clear governance frameworks that define who's responsible when something goes wrong. And honest conversations about what AI can and cannot do. The enthusiasm is already there. The infrastructure isn't.
Il Polso
- A 56-point chasm separates those who believe AI will transform medicine (86.5%) from those who actually feel well-informed about it (20.3%), raising real fears of misuse and overreliance in clinical settings.
- Younger professionals — the very cohort expected to lead adoption — show the widest knowledge deficit, with only 17.3% of those under 35 reporting high familiarity with healthcare AI.
- Institutional support is nearly invisible: only one in four respondents works where AI is actively encouraged, and fewer than one in five has access to any funding for AI training.
- Despite systemic neglect, nearly half of professionals say they would invest their own time and money to learn — a reservoir of motivation that healthcare systems are currently leaving untapped.
- Clinicians show nuanced judgment about AI's role, welcoming it for documentation and radiology while resisting its use in triage and treatment decisions — suggesting readiness for guidance, not just enthusiasm.
Across hospitals and clinics, a quiet paradox is taking shape: the healers most eager to welcome artificial intelligence into medicine are, by their own admission, among the least prepared to wield it wisely. A survey of 148 healthcare professionals from Germany and 17 other nations finds that enthusiasm for AI runs 56 percentage points ahead of actual knowledge — a gap that mirrors a broader human tendency to embrace transformation before understanding it. The moment calls not for dampened hope, but for institutions to meet the hunger their professionals are already feeling.
A survey of 148 healthcare professionals has laid bare a striking contradiction at the center of medicine's AI moment: the people most excited about artificial intelligence are, in large part, the least equipped to use it safely.
The numbers are unambiguous. While 86.5 percent of respondents believed AI would transform medical practice and three-quarters expressed genuine excitement, only 20.3 percent felt well-informed about AI in healthcare specifically. That 56-point gap between enthusiasm and knowledge is what researchers flag as the real danger — not resistance to AI, but uncritical embrace of tools that clinicians don't yet understand.
Conducted between September 2024 and January 2025, the study drew primarily from Germany, with participants across 17 additional countries. The sample skewed young and highly educated, yet even this digitally-engaged cohort showed the same pattern. Professionals under 35 were among the most enthusiastic and the least knowledgeable, with only 17.3 percent reporting high familiarity with healthcare AI.
Institutional support — the most obvious remedy — is largely absent. Only 25.7 percent work at organizations actively using or encouraging AI, and just 16.9 percent have access to institutional funding for training. Yet 45.3 percent said they would personally invest in learning, revealing a motivation that systems are failing to channel.
Barriers run deep at every level. Nearly three-quarters of respondents cited a lack of AI knowledge among healthcare decision-makers as a major obstacle. Integration difficulties, regulatory uncertainty, and privacy concerns compounded the picture. Only 2 percent reported facing no barriers at all.
Where professionals did show sophistication was in their sense of appropriate use. Translation tools, documentation support, and radiology analysis were welcomed; AI involvement in triage, treatment planning, or direct patient communication was viewed with warranted caution. Only 38.6 percent had ever used AI in clinical practice, with ChatGPT the most common tool, deployed mainly for summarization.
The researchers found that enthusiasm and knowledge functioned as entirely separate dimensions — someone could be deeply excited about AI while feeling wholly unprepared to implement it. What happens next depends on whether healthcare institutions choose to close that gap through structured literacy programs and governance frameworks, or allow well-intentioned clinicians to navigate powerful tools without the preparation they need.
A survey of 148 healthcare professionals has exposed a troubling disconnect at the heart of medicine's AI revolution: doctors and nurses are eager to embrace artificial intelligence, yet most lack the foundational knowledge to use it safely.
The numbers tell the story plainly. When asked whether AI will transform medical practice, 86.5 percent agreed. Three-quarters reported genuine excitement about AI's arrival in healthcare. But when the same professionals were asked whether they felt well-informed about AI in general, only 31.3 percent said yes. For healthcare AI specifically, the figure dropped to 20.3 percent. The gap between enthusiasm and knowledge stretched to 56 percentage points—a chasm that researchers worry could lead clinicians to misuse tools they don't fully understand.
The study, conducted between September 2024 and January 2025, drew participants primarily from Germany, with additional respondents from 17 other countries. The sample skewed toward early-career professionals: 77.7 percent were under 40 years old, and 86.5 percent held advanced degrees. These were not technophobes. Yet even among this educated, digitally-engaged group, the knowledge-enthusiasm gap persisted across age groups and geographies. Younger professionals showed the widest gap, with only 17.3 percent of those under 35 reporting high knowledge of healthcare AI, despite enthusiasm levels remaining uniformly high.
Institutional support emerged as a critical factor—but one that remains largely absent. Only 25.7 percent of respondents worked at institutions actively using or encouraging AI. Another 34.5 percent said AI wasn't even part of organizational conversations. When it came to funding, the picture was bleaker: just 16.9 percent had access to institutional financial support for AI training. Yet 45.3 percent of professionals said they would personally invest time and resources in learning, revealing a hunger for knowledge that institutions are failing to feed.
The barriers to adoption ran deep. At the healthcare system level, 72.3 percent cited lack of knowledge among decision-makers as a major obstacle. Integration challenges with existing systems troubled 64.2 percent. Regulatory uncertainty and privacy concerns each affected roughly half the respondents. At the institutional level, similar patterns emerged, with knowledge deficits among leadership cited by 62.8 percent. Only 2 percent of respondents reported facing no barriers at all.
When asked about specific AI applications, professionals showed sophisticated judgment about what belonged in clinical care and what didn't. Translation tools, documentation support, and radiology analysis ranked highest in perceived usefulness. Applications involving treatment planning, emergency triage, or direct patient communication scored lower—reflecting appropriate caution about replacing human judgment in high-stakes decisions. Notably, only 38.6 percent reported any prior experience actually using AI in clinical practice. Among those who had, ChatGPT was the most common tool, used mainly for summarization and document drafting.
The researchers found strong internal consistency in their survey measures, with enthusiasm items clustering tightly together and knowledge items correlating strongly with each other. But knowledge and enthusiasm showed almost no relationship—they operated as separate systems in the minds of these professionals. Someone could be thrilled about AI's potential while feeling completely unprepared to implement it.
What comes next will determine whether this enthusiasm becomes a force for better medicine or a liability. Healthcare systems face an urgent choice: invest in structured AI literacy programs, governance frameworks, and hands-on training that translates professional motivation into genuine competency, or watch as well-meaning clinicians deploy powerful tools they don't fully understand.
Citazioni salienti
Healthcare professionals demonstrate good understanding of appropriate AI applications and professional responsibilities, with 82.3% believing medical professionals should be engaged in AI development and 85.1% recognizing AI's potential to support clinical care.— Study findings
The mismatch between enthusiasm and readiness may contribute to inconsistent adoption, underutilization or inappropriate reliance on AI applications, including automation bias where clinicians defer to algorithmic outputs even when lacking foundational understanding of AI's limitations.— Study discussion
Vuoi la storia completa?
Leggi l'originale su PLOS (Public Library of Science) ↗
Health & Medicine