AI in clinical settings is creating both alignment and tension

Rapid increases in AI use clearly show that patients and clinicians believe strongly in the potential of the technology to improve their experiences. However, patients and clinicians come into their relationship from different perspectives, and have distinct priorities and preferences even as they work toward the same goal of better outcomes and smoother experiences.

That leaves healthcare leaders with the challenge of supporting a wide range of AI-driven use cases that are both overlapping and, at times, in tension with one another.

As a result, health systems may find themselves pulled in multiple directions when determining where and how to invest in AI tools, especially in an environment of limited resources and competing demands.

At a high level, there is agreement on the goal. Roughly 61% of clinicians believe AI can help clinicians spend more time on patient care, and 74% of patients feel that generative AI can be an efficient way for clinicians to seek information related to their care, pointing to a joint desire to use AI to reduce friction and improve the overall care experience. But beyond that headline, the definition of “improvement” begins to diverge.

 

AI is being adopted across healthcare, but not always for the same reasons— creating new tensions between operational efficiency and patient expectations.

Is GenAI making it possible for you to spend more time on patient care?

Three donut charts showing 61% of total respondents, doctors, and nurses say generative AI helps them spend more time on patient care.

  • Graphic description

    This graphic presents three donut charts comparing responses from total respondents, doctors, and nurses to the question, “Is GenAI making it possible for you to spend more time on patient care?” In all three groups, 61% of respondents answered “Yes,” indicating a majority believe generative AI enables more time for patient care. Among total respondents, 23% answered “No,” and 16% said “Don’t know.” Among doctors, 24% answered “No,” and 15% selected “Don’t know.” Among nurses, 23% answered “No,” and 16% said “Don’t know.” The visual shows consistency across all groups, with similar proportions reporting positive, negative, and uncertain responses, highlighting broad agreement that generative AI contributes to increased time for patient-focused activities.

Operational AI in healthcare is solving workflow challenges with mixed patient buy-in

For clinicians, the value of technology-driven improvements is tied to eliminating operational pain points, reducing administrative burdens, and gaining cognitive support. Over the next three years, clinical respondents envision continued challenges with cost containment pressures (79%), time spent on administrative tasks such as prior authorizations (79%), and staffing shortages or retention issues (74%), all of which represent prime use cases for AI-assisted relief.

Patients are less focused on these internal pressures – and less uniformly comfortable with how AI is applied to address them.

For example, 7 in 10 patients overall are comfortable with AI being used as part of obtaining prior authorizations. While 63% of patients approve of clinicians using AI to answer their emails, they’re marginally more amenable (73%) to one of the most popular uses of AI in the workflow: using ambient listening for transcribing conversations and generating clinical documentation.

Healthcare professional in a white coat using a desktop computer at a clinical workstation, with anatomical diagrams and office supplies visible in the background.

Patient trust in healthcare AI varies across demographics

These differences become even more pronounced across demographic groups. It may not be surprising that patients aged 55 and up are generally less comfortable than younger individuals with these use cases. But the data also found a clear split by gender, regardless of age. Male patients expressed more comfort with AI use than female patients across the board, with female respondents showing markedly higher reluctance for AI to take part in diagnosing conditions, analyzing test results, and prescribing treatments.

This is a clear signal that clinicians and patients may not yet be fully aligned around all of the health system’s most high-priority operational use cases. Healthcare leaders will need to take a considered and patient-centered approach to identifying where AI can provide the most value while still respecting patient preferences as AI reshapes the clinical experience.

Recognizing what patients do want to see from AI will be an important step in the right direction. As evidenced by their widespread use of consumer-grade AI tools to gather information on their own, patients want access, clarity, speed, and self-navigation capabilities that will help them have more informed conversations with clinicians and make more personalized decisions.

How comfortable are you with AI being used to do the following autonomously, i.e., completely on its own?

Bar chart showing patient comfort with autonomous AI in healthcare tasks, with higher comfort for administrative tasks and lower comfort for diagnosis and treatment decisions.

Percent responding Totally Comfortable

  • Graphic description

    This horizontal bar chart shows the percentage of total patients, male patients, and female patients who report being “totally comfortable” with AI being used autonomously across a range of healthcare tasks. Comfort levels are highest for administrative tasks: appointment scheduling (79% total, 84% male, 74% female), notetaking and documentation (73%, 78%, 69%), and writing emails to insurers for prior authorization (71%, 78%, 64%). Moderate comfort is reported for renewing or prescribing medication (69%, 76%, 62%) and answering emails from patients (63%, 76%, 50%). Lower comfort levels are seen for more complex clinical tasks, including interpreting imaging such as X-rays (59%, 71%, 47%), making insurance coverage determinations (57%, 68%, 47%), prescribing medication based on a new diagnosis (52%, 63%, 41%), analyzing test results and ordering treatment (52%, 62%, 42%), and making a diagnosis and recommending treatment (48%, 61%, 36%). Across all tasks, male patients consistently report higher comfort levels than female patients. Overall, the chart shows that patients are more comfortable with AI handling administrative functions than making complex clinical decisions.

Explore The Future Ready Healthcare Survey Report

Take the pulse of where AI is working in the clinical setting and where there is opportunity to do more as we better understand the unique perceptions of doctors, nurses, and patients around agentic and generative AI.

Report Home
Report Home
Key Survey Findings
Key Survey Findings
Looking Ahead
Looking Ahead
Abstract image of flowing, multicolored light waves in blue, purple, and red tones against a dark background, suggesting motion and digital data movement.
Back To Top