When AI en­ters the ho­spi­tal: Who sha­pes the fu­tu­re of he­alt­hca­re work?

But what happens to professional judgment when technology begins to indicate where abnormalities are most likely to be found? And how does work change when AI becomes a permanent participant in clinical decision-making?

These are some of the questions explored by researchers in the Nordic research project AI-PROCARE. Bringing together researchers from across the Nordic region, the project investigates how AI is procured and implemented in healthcare organizations, and how these technologies shape working conditions, professional practice, and everyday healthcare work.

The growing interest in AI within healthcare is often driven by urgent challenges. Across many countries, healthcare systems face staff shortages, increasing demand, and pressure to improve efficiency. In this context, AI is frequently presented as part of the solution. Discussions tend to focus on innovation, automation, and productivity gains.

Much less attention is paid to a different set of questions: How can healthcare organizations procure AI systems in ways that support the work environment? How does AI affect the people who work with it? And how does it reshape professional expertise, workplace autonomy, and the organization of care?

Looking Beyond the Technology

A sociotechnical perspective forms the foundation of AI-PROCARE. Rather than viewing technology as something that can simply be installed within an organization, the project treats technology and organizations as mutually shaping one another.

AI systems are influenced by local workflows, professional norms, management structures, and workplace cultures. At the same time, the systems themselves influence how work is organized, how decisions are made, and how professionals interact with one another.

This perspective places particular emphasis on procurement. Many of the most consequential decisions about AI are made long before a system reaches a hospital ward or clinical department.

What requirements are specified when a system is purchased? How flexible is the technology? Can employees adapt it to fit local practices? Does it support professional judgment, or does it gradually reshape and constrain it?

These questions are often overlooked in discussions about digital transformation, yet they may ultimately determine whether AI becomes a useful support tool or a source of new challenges.

Between Political Ambitions and Organizational Realities

Healthcare organizations are currently navigating a complex landscape of competing pressures.

In some regions, acute workforce shortages create strong incentives to adopt AI solutions quickly. Elsewhere, organizations prefer to see stronger evidence of effectiveness before implementing new systems. Resources, technological experience, and organizational needs also vary significantly across hospitals, departments, and professional groups.

A technology introduced into an intensive care unit, where sophisticated digital tools have long been part of daily work, may be received very differently from the same technology introduced elsewhere in the healthcare system.

This highlights an important insight from sociotechnical research: technologies do not produce identical outcomes everywhere. Their effects depend on the context in which they are introduced and the people who use them.

When Automation Creates New Work

One of the project's key interests lies in examining the less visible consequences of AI implementation.

Technologies are often promoted as tools that reduce workload. Yet research increasingly shows that automation can also create new forms of work.

Speech-to-text systems provide a useful example. Designed to reduce administrative burdens by automatically transcribing clinical documentation, these systems can save time when they function well. However, if they misinterpret medical terminology or produce inaccurate transcripts, healthcare professionals must spend additional time reviewing, correcting, and validating the output.

Rather than eliminating work, the technology may shift it into new forms.

Similar questions arise around professional autonomy and expertise.

When AI systems become part of diagnostic processes or clinical decision support, healthcare professionals may increasingly encounter AI-generated assessments before making their own evaluations. Over time, this raises important questions about how professional judgment develops and evolves.

If radiologists, physicians, or nurses consistently see AI recommendations first, how might this influence their own interpretations? How can organizations strike a balance between technological support and professional independence?

These are not questions about whether AI is inherently beneficial or harmful. Instead, they concern how technologies are integrated into professional practice and how organizations can ensure that AI strengthens rather than undermines expertise.

Learning from Earlier Technologies

The researchers do not approach AI from a position of technological skepticism. On the contrary, the project highlights significant opportunities for AI to improve healthcare when implementation is guided by attention to working conditions, flexibility, and local adaptation.

Previous research by members of the team offers a useful illustration.

In an earlier study of service robots in hospitals, researchers observed how employees gradually adapted the technology to fit their own work practices. Staff gave robots names, incorporated them into daily routines, and transformed them from unfamiliar equipment into collaborative elements of everyday work.

These observations demonstrated that successful implementation is rarely a one-way process in which technology simply changes organizations. Instead, employees actively shape how technologies are used and what roles they ultimately play.

AI-PROCARE builds on these insights by asking similar questions about artificial intelligence. How do new technologies become part of everyday working life? Under what conditions are they experienced as helpful rather than burdensome? And how can organizations create environments where technology supports both the quality of work and the people performing it?

Shaping the Future of Healthcare Work

As AI becomes more deeply embedded in healthcare, the conversation extends beyond technology itself.

AI increasingly becomes part of the relationships, routines, decisions, and professional practices that define everyday healthcare work. Decisions about procurement, implementation, and governance are therefore also decisions about the future organization of care.

Who gets to influence how AI is used? How much flexibility should healthcare professionals have in adapting technologies to local needs? And how can organizations ensure that digital transformation supports both efficiency and professional expertise?

These questions sit at the heart of AI-PROCARE.

Ultimately, the project reminds us that AI is not simply introduced into healthcare systems. It becomes woven into the social and organizational fabric of healthcare itself. Understanding that process may be just as important as understanding the technology.