2025 Future Health Index: Building Trust in Healthcare AI
A Report on Building Trust in Healthcare AI
Commissioned by Philips
Foreword
The healthcare system is at a crossroads. Workforce shortages, rising costs, and system inefficiencies are pushing healthcare to its limits, and patients are feeling the impact. The 2025 Future Health Index, the 10th edition and largest survey ever conducted, reveals that more than half of surveyed countries report patients waiting nearly two months for specialist care. Without decisive action, a shortage of 10 million healthcare professionals is projected by 2030, leaving millions unable to receive essential care in a timely manner.
Artificial intelligence (AI) is emerging as a powerful catalyst and opportunity to meet the growing demand for healthcare services driven by an aging population. Considering the rapid advancements in science and technology over the past five years, and anticipating the next five, AI is expected to automate a significant portion of the 'silent' administrative tasks performed by healthcare professionals, freeing up their time and significantly enhancing their clinical capabilities.
Surveyed healthcare professionals recognize AI's potential to help them reclaim time lost to administrative duties, diagnose diseases more accurately, reduce preventable hospital readmissions, and improve treatment outcomes. Further research suggests that widespread adoption of AI could save between $200 billion and $360 billion annually in healthcare costs in the United States alone.
While AI is advancing rapidly, public trust is lagging. Although most healthcare professionals surveyed in the 2025 Future Health Index are optimistic about AI's ability to improve healthcare services, many patients remain skeptical, especially when their health is at stake. Furthermore, despite optimistic outlooks, a majority of healthcare professionals still express significant concerns about bias and accountability. Without trust, AI's full potential in healthcare cannot be realized.
Building trust requires a responsible, human-centered approach that places collaboration at the heart of AI innovation. AI should enhance, not erode, the trust between patients and healthcare professionals. It must deliver tangible benefits, be underpinned by robust safeguards, and operate within a clear and consistent regulatory framework. Only then can AI gain the trust needed to drive meaningful change in healthcare.
This means accelerating AI innovation in the right direction to save more lives, not slowing it down. It requires collaboration across disciplines, institutions, and borders. This report provides key insights to foster that collaboration. We call on healthcare leaders worldwide to act on these insights, ensuring technology and trust advance hand-in-hand to create a future where better healthcare is accessible to more people.
Shez Partovi
Chief Innovation Officer (CIO)
Head of Global Business Leader, Health Informatics
Carla Goulart Peron
Chief Medical Officer (CMO)
"As AI transforms healthcare, trust and innovation must go hand-in-hand to deliver life-saving solutions to more people, faster, and with the right safeguards."
Survey Overview
This survey is the largest of its kind, analyzing the priorities and perspectives of healthcare professionals and patients. The 10th annual Future Health Index report explores how AI can help healthcare professionals provide better care for more people. The report focuses on key challenges affecting healthcare professionals today, examining their views on the rise of AI and identifying critical gaps that need to be addressed to build greater trust in AI-driven patient care. It also assesses the patient perspective on the use of healthcare AI and opportunities to strengthen trust in technological advancements.
This year's Future Health Index report is based on original quantitative research conducted among 1,900 healthcare professionals and over 16,000 patients across 16 countries.
Description of map: A world map highlights 16 countries where the survey was conducted, including North America (USA, Canada), Europe (UK, Netherlands, Germany, France, Spain), South America (Brazil), Middle East (Saudi Arabia), Asia (China, Japan, South Korea), and Oceania (Australia), South Africa.
Key Statistics:
- 16 Countries surveyed
- 1,900+ Healthcare Professionals
- 16,000+ Patients
Chapter 1: AI-Driven Healthcare Innovation
Longer Treatment Waits Burdening Patients
Like many countries globally, South Korea faces increasing pressure from an aging population, a rise in chronic diseases, and a growing gap between patient demand and healthcare supply. With a doctor-to-patient ratio of just 2.6 per 1,000 people, among the lowest in developed nations, urgent reform is needed to address the projected large shortage of physicians.
The survey indicates this issue significantly impacts patients, with over one-third reporting long wait times for appointments. More than half of patients (53%) have experienced waiting for specialist care, with an average wait time of 40 days, which is lower than the global average of 70 days. Treatment delays have real consequences: 1 in 5 Korean patients (21%) reported their health worsened due to delayed access to doctors, and 14% have experienced hospitalization because of long wait times.
Description of bar chart: The bar chart displays the average number of days patients wait for specialist appointments across various countries. Countries shown include Australia (131 days), Brazil (98), Canada (99), China (93), France (18), Germany (40), India (19), Indonesia (22), Japan (63), Netherlands (51), Saudi Arabia (59), Spain (128), South Africa (109), South Korea (40), UK (33), and USA (40). A statistic highlights that 53% of Korean patients have experienced waiting for specialist care. A separate indicator shows the global average wait time is 70 days.
Healthcare Professionals Prioritize Patient Care Despite Time Constraints
Despite facing system-level issues that consume valuable time, healthcare professionals remain dedicated to their mission. Data inefficiencies often lead to professionals wasting precious time and energy that could be better spent on patient care. Nearly all healthcare professionals (91%) report spending clinical time dealing with incomplete or inaccessible patient data, a significantly higher percentage than the global average of 77%. Half of these professionals report losing over 45 minutes per shift, totaling approximately 23 days of lost work time annually per professional. This underscores the urgent need for AI and digital technologies to streamline data management, allowing professionals more time with patients.
Description of infographic: A statistic shows that 91% of healthcare professionals experience issues due to incomplete or inaccessible patient data. Another statistic indicates that 51% of healthcare professionals lose over 45 minutes per shift due to these issues. A calculation estimates this loss to be over 4 weeks annually per professional.
Administrative Tasks Consume Time, With Little Improvement
The burden of administrative tasks is growing for healthcare professionals struggling to access necessary data. A survey reveals that 30% of Korean healthcare professionals spend more time on administrative tasks than five years ago, while spending less time with patients. Despite this, more than half of healthcare professionals still find meaning and reward in comforting patients and supporting their recovery.
Description of bar chart: The bar chart illustrates the trend of administrative tasks for Korean healthcare professionals. 29% report spending less time with patients and more on administrative tasks. 44% report spending equal time on patient care and administrative tasks. 27% report spending more time with patients and less on administrative tasks.
AI to Provide Better Healthcare for More People
Healthcare professionals overwhelmingly believe that implementing AI correctly will allow them to see more patients and reduce wait times, reclaiming valuable time. They anticipate AI will enable more accurate and timely medical interventions, allowing professionals to spend more face-to-face time with patients. This perception aligns closely with that of healthcare workers in the Asia-Pacific region.
Healthcare professionals recognize the urgency of addressing issues of delayed care and inefficiency, warning of the opportunity cost if AI is not adopted more quickly. They fear that delayed adoption could lead to burnout among healthcare staff (41%), a decline in the quality of patient care (40%), and increased patient backlogs (39%). They also worry about being unable to provide cutting-edge treatments (38%).
Description of bar charts: These charts illustrate the perceived positive impacts of AI in healthcare. Under 'Patient Access & Experience,' key findings include: 92% believe AI will expand patient access, 91% expect reduced wait times, 86% anticipate better patient triage, and 85% foresee increased face-to-face time with patients. Under 'Clinical Excellence & Innovation,' 89% believe AI will enable timely interventions, and 84% expect improved access to clinical research. Under 'Operational Efficiency & Workflow Optimization,' 85% expect automation of repetitive tasks and improved patient throughput, while 80% expect reduced treatment times.
AI-Driven Innovation in Preventive Healthcare
While AI is already transforming healthcare services, its greatest potential lies in preventing the need for certain types of treatment. As chronic diseases and their associated costs rise, South Korea's government and healthcare system are shifting from a reactive to a proactive approach, focusing more on preventive care.
Survey results indicate that Korean healthcare professionals believe AI-powered predictive analytics and remote patient monitoring can enable timely interventions, saving lives by preventing hospitalizations for some patients. They are embracing this shift.
Description of donut charts: These charts highlight the perceived benefits of AI in preventive healthcare. 90% believe AI can save lives through early intervention. 86% believe AI can reduce the need for acute or emergency medical treatment. 84% believe AI can reduce hospital admissions.
Chapter 2: The Healthcare AI Trust Gap
Building trust is crucial for the widespread adoption of AI in healthcare. While many patients and healthcare professionals have positive expectations for AI, concerns are being voiced about patients' trust lagging behind that of healthcare professionals. How large is this trust gap, and how can it be narrowed?
Healthcare Professionals Are More Optimistic About AI Than Patients
Despite AI's rapid development and potential, its adoption and impact in healthcare depend not only on technological advancements but also on the trust and acceptance of both healthcare professionals and the patients they serve. Survey results show that the majority of Korean healthcare professionals are significantly more optimistic about AI's potential to improve healthcare services than patients are, with a gap of 26 percentage points. This presents a significant challenge for healthcare leaders, policymakers, and industry stakeholders to maximize AI's benefits without undermining patient trust and acceptance.
Description of donut charts: The first chart shows that 86% of Korean healthcare professionals believe AI will improve patient care, compared to 60% of Korean patients. The second chart shows that 85% of healthcare professionals in the APAC region believe AI will improve healthcare services, compared to 59% of patients in the APAC region. Globally, 79% of professionals and 59% of patients hold similar positive views.
The AI Trust Gap Persists Between Healthcare Professionals and Patients
The AI trust gap between healthcare professionals and patients exists even when considering specific AI applications for patient care. While healthcare professionals generally show high trust in AI use, patients tend to be more cautious. Korean patients show higher trust compared to the global average in applying AI for clinical tasks such as emergency patient triage, treatment planning, and medical record keeping.
Description of bar chart: This chart illustrates the AI trust gap between healthcare professionals and patients across various AI applications. For 'Medical Record Keeping,' professionals' trust is 92% vs. patients' 70%, a gap of -22. For 'Personalized Treatment Planning,' it's 92% vs. 71% (-21 gap). For 'Diagnostic Support,' it's 87% vs. 71% (-16 gap). For 'Clinical Decision Support,' it's 85% vs. 70% (-15 gap). For 'Patient Check-in,' it's 85% vs. 73% (-12 gap). For 'Appointment Scheduling,' it's 84% vs. 76% (-8 gap). For 'Emergency Patient Triage,' it's 84% vs. 67% (-17 gap). The chart notes that minor discrepancies may occur due to rounding.
Some Patients Fear Reduced Face Time with AI, But Most Welcome Technology If Benefits Are Clear
While some patients worry about AI leading to reduced face-to-face consultations, the majority welcome technological advancements if they offer clear benefits. Patients are accustomed to brief consultations with doctors, with the report indicating an average of just 7 minutes per visit, and two-thirds of patients having consultations under 5 minutes. This brevity leaves little opportunity for patients to ask questions about their condition.
Concerns about technology further reducing doctor-patient interaction are understandable. However, the survey results show that less than half of patients express this concern. Furthermore, over two-thirds of respondents indicated they would welcome increased adoption of technology if it improves healthcare access and benefits patients like them.
Description of bar charts: The first set of bars shows the percentage of patients who welcome more technology if it improves healthcare services. For example, 76% in Australia, 90% in Brazil, 85% in Canada, 84% in China, 82% in France, 80% in Germany, 72% in India, 81% in Indonesia, 76% in Japan, 71% in Netherlands, 67% in Saudi Arabia, 62% in Spain, 56% in South Africa, 56% in South Korea, 49% in the UK, and 49% in the USA. The second set of bars shows the percentage of patients who worry about reduced face time with doctors due to increased technology adoption. For example, 46% in Australia, 53% in Brazil, 51% in Canada, 52% in China, 43% in France, etc.
New Technologies' Utility Still Falls Short of Healthcare Professionals' Expectations
While Korean healthcare professionals are generally optimistic about AI's potential, they are skeptical about the utility of new digital health technologies, including AI, in routine clinical practice. Despite 84% of healthcare professionals participating in the development of these technologies, only 46% feel they are designed to meet their specific needs. This finding is consistent with research conducted in the Asia-Pacific region.
This indicates a significant gap in translating clinical needs into practical solutions that streamline daily workflows. The report highlights that professionals want technologies that support their daily tasks and improve patient outcomes, but current offerings often do not fully meet these expectations.
Description of infographics: One infographic shows that 8 out of 10 healthcare professionals are actively involved in developing new technologies within their organizations. Another infographic indicates that 4 out of 10 professionals believe new technologies are designed to meet their needs.
Healthcare Professionals Are Unsure About Accountability for AI
A persistent concern among healthcare professionals is: 'Who is responsible if an AI system makes a diagnostic or treatment error?' Issues such as AI 'hallucinations' (generating inaccurate or nonsensical outputs) that undermine accuracy and reliability are prompting professionals to call for updated accountability regulations. This could include holding a broader range of stakeholders, including manufacturers, more responsible for medical errors caused by AI.
Survey results indicate that approximately three-quarters of Korean healthcare professionals are concerned or unsure about legal accountability if an AI system makes an error in diagnosis or treatment. This suggests a significant barrier to AI adoption.
Description of donut chart: A statistic shows that 74% of healthcare professionals are concerned or unsure about legal accountability in case of AI errors.
Chapter 3: Bridging the Trust Gap
What is needed to strengthen trust in AI for both patients and healthcare professionals? The survey results provide a clear direction for more effective and trustworthy AI adoption in healthcare, ultimately contributing to improved patient outcomes and the overall patient experience.
Patients Expect Key Benefits from AI
To understand what makes patients more receptive to AI in healthcare, direct feedback was sought. Beyond the potential benefits for their treatment experience, patients are more open to AI when it offers faster, more accurate, efficient, and cost-effective care. Survey results indicate that their perspectives align with those of patients in other Asia-Pacific regions.
Patients are more open to AI when it leads to more personalized interactions. This can alleviate concerns about reduced face-to-face time with doctors. If implemented correctly, AI can actually lead to more personalized healthcare services.
Description of bar charts: These charts show patient expectations for AI benefits. In South Korea, 50% expect AI to reduce errors, 43% expect lower costs, 40% expect improved health outcomes, 34% expect faster access to doctors, 31% expect doctors to spend less time on administrative tasks, and 25% expect doctors to spend more time on patient consultations. Corresponding figures for the APAC region are 41%, 39%, 40%, 42%, 30%, and 32% respectively.
Patients Who Are Familiar with AI Want Stronger Guarantees of Safety and Security
Patients who are more knowledgeable about AI tend to feel more comfortable with its use. However, these patients also demand stronger guarantees. Specifically, they want to know if the technology was developed within their country's healthcare system, how it will be used in their treatment process, and whether professionals will review AI-generated recommendations.
Survey findings suggest that familiarity with AI alone does not eliminate concerns; it can even amplify them. Patients knowledgeable about AI's potential benefits feel more comfortable but are also more aware of its risks and limitations, thus demanding greater transparency and control.
Description of bar charts: The first chart shows that 45% of patients who are familiar with AI want to understand how AI is used in healthcare, compared to 32% of those less familiar. The second chart shows that 41% of familiar patients want healthcare professionals to always review AI suggestions, compared to 32% of less familiar patients. The third chart shows that 30% of familiar patients want the technology to be developed within their national healthcare system, compared to 17% of less familiar patients.
Information from Healthcare Professionals Is Key to Gaining Patient Trust
When it comes to building trust, most patients prefer to receive information from their doctors and the healthcare system. Patients consider the source of information very important when deciding whether to trust it.
Description of bubble charts: These charts show patient comfort levels with AI-based treatments based on the source of information. Doctors are the most trusted source (84% comfort in Korea, 86% in APAC). The healthcare system is also highly trusted (84% in Korea, 83% in APAC). Nurses are trusted by 80% in Korea and 80% in APAC. Friends or family are trusted by 78% in Korea and 75% in APAC. Technology developers are trusted by 77% in Korea and 73% in APAC. News sources are trusted by 74% in Korea and 70% in APAC. Social media is the least trusted source (58% in Korea, 57% in APAC).
Healthcare Professionals Demand Clear Guidelines to Build AI Trust
What is needed to further enhance the trust of clinical healthcare professionals in AI? Survey results highlight clear guidelines on AI use and limitations, reliable IT/support desk services, clarity on legal responsibility for AI use, and scientific evidence of its effectiveness as key factors. Healthcare professionals also demand continuous monitoring and evaluation of AI systems, increased transparency, and assurance of ongoing effectiveness.
A higher percentage than the global average (10 percentage points higher) desire job security. This suggests that while healthcare professionals see AI primarily as a valuable tool for enhancing their skills, many still perceive it as a threat to their careers. However, the overall sentiment is clear: with appropriate safeguards in place, the future of AI in healthcare is bright, promising improvements in healthcare services, increased efficiency, and better patient outcomes.
Description of list: Key elements required by healthcare professionals to build AI trust include: 39% want clear guidelines on AI use and limitations. 39% want reliable IT/support desk services. 36% want clear regulations on legal responsibility for AI use. 36% want scientifically proven effectiveness. 33% want assurance of job security. 31% want continuous monitoring and evaluation of AI systems. 29% want transparency on the basis of AI recommendations. 26% want data security assurances. 25% want mitigation of data bias and quality issues.
Workflow Integration Challenges Hinder AI Adoption
One of the primary reasons AI and digital health applications fail to integrate into actual clinical settings is the lack of workflow integration. This is a major barrier to AI adoption, also reflected in healthcare professionals' concerns about IT support. For example, radiologists often manage multiple software applications across various screens in high-pressure, complex situations. If an AI tool requires managing yet another application, it could increase workload rather than improve efficiency.
To truly support healthcare professionals and enhance productivity, AI algorithms must integrate seamlessly with existing systems, minimizing the need for task switching. This ensures AI acts as a helpful tool rather than an additional burden.
Recommendations
How can we build trust in AI for both patients and healthcare professionals?
- Prioritize Human-Centric AI Design: AI should be designed with the needs of patients and healthcare professionals at its core. Engaging relevant stakeholders early and collaborating continuously is essential for building trust and acceptance. Solutions must seamlessly support patient care routines and integrate with IT infrastructure, providing a frictionless experience for healthcare professionals and improving patient outcomes.
- Enhance Human-AI Collaboration: AI's true potential lies in augmenting the capabilities of healthcare professionals and empowering patients and caregivers to manage their health and well-being. While AI agents can perform specific tasks autonomously, human oversight is critical in situations where health is at risk. Healthcare professionals play a pivotal role in building patient trust through transparent communication about AI's role, supported by comprehensive training from the early stages.
- Demonstrate Efficacy and Fairness: Both healthcare professionals and patients want assurance that AI will function as intended, and regulators require evidence of safety and performance standards. Consistent performance across relevant patient populations and clinical settings is essential, along with measures to prevent biased outcomes. Development and validation should use representative and high-quality data to mitigate bias and ensure equitable results for all patients.
- Foster Innovation Based on Clear Guidelines: To rapidly provide life-saving AI to patients, regulations must balance the pace of innovation with safeguards for patient protection and trust-building. Global harmonization of regulatory frameworks can reduce complexity and accelerate the adoption of AI innovations without compromising patient safety. Approaches like regulatory sandboxes can enable responsible AI development and monitoring, while promoting consistent application of medical device regulations.
- Build Strong Cross-Sector Partnerships: No single entity can solve the challenges in healthcare alone. Close collaboration among all ecosystem stakeholders—healthcare providers and professionals, patient advocacy groups, payers, regulators, and the health tech industry—is crucial to drive innovation and create solutions that meet stakeholder needs and build trust. Most importantly, enhancing the health and well-being of patients and healthcare professionals must be the shared goal, aligning objectives and incentives, particularly reward models.
Appendices
Survey Methodology
Two quantitative surveys were conducted by Accenture Song, a global leader in technology-driven creative services, using Computer-Assisted Web Interviewing (CAWI). The surveys were administered across 16 countries: Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Japan, Netherlands, Saudi Arabia, Spain, South Africa, South Korea, the UK, and the USA. The surveys were conducted from December 2024 to April 2025.
Survey 1: 1,926 Healthcare Professionals (15-minute online survey). Respondents included physicians (including surgeons), nurses, and physician assistants, working in both private and public healthcare systems across various specialties.
Survey 2: 16,144 Patients aged 18+ (10-minute online survey). Respondents were representative of their respective countries in terms of age and gender. 99% of respondents had experienced a doctor's visit within the past two years.
Note: While two separate surveys were conducted, the data is referenced as 'the survey' for reporting convenience. Data reflects mainland China only; Taiwan and Hong Kong are not included.
Description of tables: The tables detail the sample sizes and margin of error for the surveys, both with and without weighting. Weighting is used to adjust sample data to more accurately represent the larger population, ensuring representativeness and improving accuracy by correcting for market imbalances. The data indicates that weighted samples are used to provide representative insights across diverse markets.
Glossary
This section provides definitions for key terms used throughout the report, covering concepts such as Artificial Intelligence (AI), AI Algorithms, AI Hallucinations, Data, Data Bias, Digital Health Technology, Generative AI, Healthcare Professionals, Hospitals, Machine Learning, Predictive Analytics, Remote Patient Monitoring, Workflow, and more. These definitions help ensure a clear understanding of the technical and medical terminology discussed.
