Nurses dissatisfied with AI implementation, feel 'watched' by AI

Impact of AI use on nursing workflows

Most nurses (77%) use AI within EHRs and clinician decision support workflows, the survey shows. Many nurses also use AI-driven predictive risk alerts (66%), AI staffing and acuity algorithms (59%) and ambient AI for documentation (51%).

Still, many nurses are dissatisfied with AI implementation in their workplaces, with 63% stating that AI tools have added new tasks to their workloads without removing old ones. Another 57% said they have to produce extra documentation to explain exceptions, and 55% said they spend more time validating or correcting AI outputs.

Notably, 52% of nurses reported changing the sequence of the care workflow to meet clinical targets.

AI has also led to breakdowns in communication among nursing staff and leadership, the survey reveals. More than half of survey respondents (58%) report spending more time explaining their decisions to managers or quality, IT and compliance staff. Around 55% said that AI dashboards replace direct discussions of clinical context.

"Experienced nurses develop personal sequencing habits, cognitive checks and communication routines that help them manage competing priorities," the report states. "AI systems can disrupt these habits when they prescribe a fixed order, judge timestamp deviation, or surface a metric without the patient-level context that made the deviation clinically appropriate."

Perception of surveillance by AI

Compounding workflow disruptions is that most nurses feel "watched" or tracked by AI. About 71% said that individual activities can be traced to a nurse, and 51% of nurses said that AI-derived data is used for performance reviews or coaching.

However, several nursing tasks cannot be easily measured, the report notes. For example, nurses may spend time during their workday helping another nurse, mentoring staff or resolving a unit problem, which can be challenging to record.

In fact, 59% of nurses said that teamwork is invisible in individual metrics, and 56% said managers treat AI output as objective despite missing context. Only 38% feel safe overriding with AI.

As a result, nearly half of nurses (49%) use workarounds to reduce alerts, and 33% say they are less willing to report a near miss.

Overall, nurses do not feel that leadership has a handle on how AI impacts nursing workflows, with 69% saying that executives do not understand the bedside disruption the technology causes. A majority of nurses (71%) also said leaders emphasize AI return on investment, throughput and adoption over its impact on nurses.  

Some nurses are, therefore, looking for workplaces free of heavy AI use. The survey reveals 21% are very likely and 27% are somewhat likely to look for less AI-intensive roles until the technology matures further.

"Hospitals can show a successful AI rollout on a dashboard while nurses experience a more complicated shift," said Doug Brown, founder of Black Book Research, in the press release. "The test is not whether a system generates more information or completes one task faster. It is whether nurses have less total work, clearer communication and more time for patients."

Anuja Vaidya has covered the healthcare industry since 2012. She currently covers healthcare IT and innovation, including artificial intelligence, digital healthcare, EHRs and interoperability.