List of Jobs Most Difficult for AI to Replace, Nurses Top the List
Artificial intelligence (AI) is increasingly taking over various jobs, from writing and creating presentations to analysing data. However, not all professions face the same risk. Data from JobZone Risk combined with employment prospects from the U.S. Bureau of Labor Statistics (BLS) shows that only four of the 20 most in-demand jobs in the United States have an AI resilience score above 60. This means that jobs which are truly difficult for AI to replace are still rare. Most remain potentially subject to change as technological capabilities increase, although this does not mean they will disappear entirely.
The ranking combines two factors: the scale of labour demand and a job’s level of resilience to AI-based automation. At the top of the list is registered nurse, with an AI resilience score of 82.2. The next positions are held by home health aide (72.7), elementary school teacher (70.0), and nursing assistant/CNA (67.4). Outside of these four professions, all other jobs on the list have a score below 60.
Healthcare services dominate the ranking. Nurses, patient companions, and nursing assistants rely on abilities that are still difficult for AI to replicate, ranging from direct patient care, empathy, and clinical judgement to decision-making in constantly changing situations. The BLS also projects that demand for healthcare workers will continue to rise with the growing elderly population and healthcare needs. Meanwhile, AI is more often used to assist with medical staff’s administrative work rather than replacing direct patient interaction.
Resilience to AI is also evident in the education sector and technical jobs. Elementary school teachers scored 70, while teaching assistants recorded a score of 51.2. Jobs such as general maintenance and repair workers, with a score of 53.2, and construction workers (53.2) are also in the upper group. The commonality among these professions is the high need for on-site adaptation, creativity, decision-making, and direct human interaction. These characteristics remain a challenge for AI systems.
Conversely, some professions employing millions of people actually received lower scores. General and operations managers only scored 37.5, followed by long-distance lorry drivers (36.0), retail sales supervisors (34.7), logistics workers (29.9), and wholesale and manufacturing sales representatives at the bottom with a score of 26.1. Many tasks in these jobs relate to information processing, administration, documentation, and analysis, which are now increasingly easy for generative AI to assist with.
However, these ranking results do not indicate that professions with low scores will disappear from the job market. Instead, AI is expected to automate parts of tasks more than replace entire jobs. Therefore, technical skills alone are becoming insufficient. As AI becomes more capable of processing information, human added value is increasingly shifting towards abilities that are harder to teach machines, such as empathy, leadership, creativity, communication, and problem-solving.
Source: jawawa.id (https://jawawa.id/newsitem/list-of-jobs-most-difficult-for-ai-to-replace-nurses-top-the-list-1785826289)
Originally published by CNBC Indonesia (https://www.cnbcindonesia.com/research/20260804123026-128-756385/daftar-pekerjaan-paling-sulit-digantikan-ai-perawat-juara), translated from Indonesian