As more and more professionals are starting to rely on artificial intelligence tools in their work, could their hard-earned skills atrophy?
This possibility is increasingly worrying medical specialists, computer scientists and other workers. Seventy percent of nurses and 77% of doctors, for example, fear losing their skills because of over-reliance on AI systemsaccording to a survey of U.S. health workers released earlier this month.
Their fear could be justified. Data suggests that AI-driven “deskilling” is starting to happen in medicine, computer science, and other fields. Researchers are now discussing how to preserve important human expertise in the AI era.
On supporting science journalism
If you enjoy this article, please consider supporting our award-winning journalism by subscribe. By purchasing a subscription, you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
“Just being aware that this exists hopefully sparks some thinking about what skills people want to keep and what skills they are willing to outsource” to AI tools, says Kevin Crowston, an information scientist at Syracuse University in New York.
Spoiled by AI?
A study of Polish doctors specializing in endoscopy – the use of flexible probes to examine the inside of the human body – shows how quickly AI tools can erode human capabilities. The doctors, all of whom have performed at least 2,000 colonoscopies in their careers, were given access to an AI system that analyzes colonoscopy images in real time and reports a type of precancerous intestinal lesion called an adenoma. The tool was available to specialists on some days but not on others.
Once doctors started using it, their performance dropped significantly when the system was unavailable. In the three months preceding the introduction of the AI tool, specialists detected at least one adenoma during 28.4% of colonoscopies. In the three months since the tool was introduced, the adenoma detection rate for colonoscopies performed without AI assistance decreased to 22.4%.
The results, published last October in The Lancet Gastroenterology and Hepatologysuggest that even highly skilled professionals could deteriorate in the tasks their jobs require as they become more dependent on AI tools, says Robert Wachter, a physician at the University of California, San Francisco, author of a book on how AI tools are transforming health care. The study authors say continued exposure to such tools can cause clinicians to become “less motivated, less focused, and less accountable when making cognitive decisions without the assistance of AI.”
Co-author Yuichi Mori, a physician-researcher at the University of Oslo, says more studies are needed to confirm the phenomenon. But people who use AI tools need to be aware that they risk losing some of their skills, he adds. “There is currently no established solution to deskilling. This is expected to become a very hot research topic over the next decade.”
No lessons learned
To determine whether skills are being lost in computer science, researchers at San Francisco, Calif.-based AI company Anthropic designed a randomized controlled trial in which 52 software engineers were asked to complete a basic coding task. During the exercise, the 52 participants were able to search the web and access instructions on how to complete the task. Half of the participants were also asked to use an AI assistant.
Then, all software engineers were asked to take a quiz on what they learned from this task. Participants who had used an AI assistant performed significantly worse on the quiz than those who had not: the average score was 50% in the AI group compared to 67% in the non-AI group. AI-assisted participants performed particularly poorly on questions that required them to diagnose errors in the code, suggesting that they had failed to learn the concepts behind the code they had just produced. The study was published on the arXiv preprint server ahead of peer review.
The findings are concerning, especially for students and young professionals in the field, says Crowston, who studies how the use of generative AI tools is changing the way software developers acquire and maintain coding skills. “There is now a very strange disconnect between performance and learning,” he says. “People can achieve a pretty high level of performance because they’re essentially borrowing skills from AI, but they’re not developing those skills themselves.”
Externalization of cognition
Other technologies have made certain skills obsolete in the past, notes Tapani Rinta-Kahila, an information systems researcher at the Hanken School of Economics in Helsinki. For example, GPS navigation systems have eroded people’s navigation skills. Generative AI tools, however, are “the first technology that automates various cognitive faculties around thinking and interpretation, which have long been considered unique human skills.”
Rinta-Kahila’s own work reinforces these concerns. In 2018, he published a study on a group of accountants who had continuously used an automated, AI-free accounting system for over a decade. His team found that when the tool was taken away, accountants forgot how to perform many routine tasks. He predicts that AI systems will affect work in a variety of ways, as they take over basic tasks once performed by early-career professionals. “The next generations of programmers may not understand the fundamentals of coding very well if they don’t have the hands-on experience,” he says. “The same is true for many other knowledge-intensive professions, such as accounting and law. »
To avoid skills erosion due to AI, people need to be aware of the importance they place on generative AI tools, he says. They also need to understand exactly how generative AI models work and what their limitations are – and avoid trusting AI results without questioning them. “People need to manage the competing dynamics of relying on generative AI and remaining vigilant. »
This article is reproduced with permission and has been published for the first time June 18, 2026.
It’s time to defend science
If you enjoyed this article, I would like to ask for your support. Scientific American has been defending science and industry for 180 years, and we are currently experiencing perhaps the most critical moment in these two centuries of history.
I was a Scientific American subscriber since the age of 12, and it helped shape the way I see the world. SciAm always educates and delights me, and inspires a sense of respect for our vast and beautiful universe. I hope this is the case for you too.
If you subscribe to Scientific Americanyou help ensure our coverage centers on meaningful research and discoveries; that we have the resources to account for decisions that threaten laboratories across the United States; and that we support budding and working scientists at a time when the value of science itself too often goes unrecognized.
In exchange, you receive essential information, captivating podcastsbrilliant infographics, newsletters not to be missedunmissable videos, stimulating gamesand the best writings and reports from the scientific world. You can even give someone a subscription.
There has never been a more important time for us to stand up and show why science matters. I hope you will support us in this mission.