Reskill or Risk Replacement: OECD Flags AI’s Growing Impact on Health Occupations

A groundbreaking May 2025 report from the Organisation for Economic Co-operation and Development (OECD), authored by Fabio Manca and Diego Eslava, in collaboration with labour analytics firm Lightcast, provides a revealing look into how digital transformation and artificial intelligence (AI) are reshaping healthcare occupations. Drawing insights from an immense dataset of nearly 55.5 million online job postings across Canada, the United Kingdom, and the United States, the study maps the rise of digital and AI skill demands in the healthcare workforce. As health systems worldwide contend with staff shortages, ageing populations, and rising demand for services, the integration of digital and AI technologies is seen not only as inevitable but increasingly urgent.

Health Jobs Are Going Digital And Fast

The report finds that digital technologies are rapidly permeating the health sector, and demand for digital competencies is growing, especially in roles not directly involved in patient care. In the United States, postings requiring skills in Health Information Management and Electronic Health Records (EHRs) are on the rise, accompanied by increasing demand for expertise in web design and video conferencing. These trends suggest a broader shift towards virtual health delivery and remote patient services. In the United Kingdom, job postings increasingly mention Cybersecurity and Data Analysis, underscoring priorities around data protection and advanced monitoring technologies. Meanwhile, Canada shows consistent interest in Telehealth, Simulation, and Clinical Informatics, especially in non-clinical functions such as IT support and administrative coordination.

Despite these trends, digital skill needs still represent a relatively modest share of total health-related job openings. For example, digital-related job postings comprise between 2% and 6% of frontline health occupation vacancies and 6% to 13% in other health sector roles, such as hospital IT and finance. Nonetheless, the report shows that the digitalisation of health services is a clear and steady trend, particularly for positions supporting the digital infrastructure of modern health systems.

AI in Health: Quietly Transforming Without a Job Title

While digital skills are taking centre stage, the explicit demand for AI-specific skills, such as machine learning or neural networks, remains limited in job descriptions. Only about 0.2% to 0.3% of total postings in health occupations listed AI skills in 2023, although the percentage rises slightly in non-clinical roles. In the U.S. and Canada, for example, postings in health administration and research are beginning to request capabilities in AI tools, predictive modelling, and data automation. However, the use of AI in clinical settings is far more pervasive than these numbers suggest.

AI is being integrated into everyday health practices, often without being explicitly named in job ads. Radiologists may use AI-powered diagnostic systems, while nurses rely on AI-assisted patient monitoring tools. Because many of these technologies are embedded into broader platforms like EHRs or hospital management software, employers often don't specify AI as a requirement. Instead, they focus on results-driven phrases like "improved diagnostic accuracy" or "real-time decision-making support." The report suggests that this silent integration of AI is likely to grow, especially as generative AI becomes more capable of supporting cognitive and administrative tasks.

Augmentation vs. Automation: Which Jobs Are Safe?

To better understand the implications of AI and robotics, the OECD uses U.S. occupational data from O*NET to assess which healthcare roles are most susceptible to automation. Using advanced modelling via large language models like GPT-4, the study categorises jobs into four groups: low risk of automation, potential augmentation, potential automation, and high risk of automation.

Most occupations, including Registered Nurses, Family Medicine Physicians, and Physician Assistants, fall into the potential augmentation category. These are roles that benefit from technology without being replaced by it. In contrast, Pharmacy Technicians and Medical Equipment Preparers are flagged as at risk of partial automation, while Orderlies and Medical Transcriptionists face the highest risk due to the repetitive and rule-based nature of their tasks.

In total, around 32% of health employment in the U.S. is likely to experience job augmentation, while 4.3% could face partial automation. Just 0.6% of the workforce is in roles that may be fully automated. This means that the future is less about machines replacing humans and more about machines helping humans do their jobs more efficiently.

Training, Equity, and Ethics: Building a Future-Ready Workforce

The policy implications of these findings are clear and pressing. First, investment in training is critical. For roles with high augmentation potential, such as nurses and physician assistants, professional development in digital systems, telehealth platforms, and wearable technology will be key to improving outcomes and job satisfaction. The report emphasizes the need to incorporate these skills into health education and continuing training programmes.

Second, policymakers must mitigate the risks for workers in highly automatable roles. Reskilling programmes, career transition pathways, and support for learning new technologies can help avoid displacement. Equally important is addressing the digital divide, ensuring that all regions, including underserved or rural areas, have access to digital infrastructure and training opportunities.

The report highlights ethical and governance considerations. As AI becomes more embedded in clinical workflows, maintaining transparency, accountability, and data privacy will be essential. Robust data governance frameworks must accompany this technological shift, so that patients and providers alike can trust the systems they use.