Bot Docs? Not Likely: 69% of US Adults Uncomfortable Being Diagnosed by AI

From wearable health devices to virtual doctor’s appointments, healthcare has come a long way. But there’s one change people don’t want to see: doctors replaced with AI. That’s according to Salesforce’s Pulse of the Patient Snapshot, which surveyed more than 1,400 adults in the U.S. and revealed that 69% of them are uncomfortable with healthcare companies using AI to diagnose them. 

However, people haven’t ruled out AI entirely. More than half of respondents are comfortable with AI in nonclinical use cases, such as scheduling appointments and estimating medical expenses. This indicates an opportunity for healthcare organizations to focus on using AI in areas that free up physicians to spend more time with their patients.

Why it matters: According to an Athenahealth survey, more than 90% of U.S. physicians report feeling burned out on a “regular basis,” especially by overwhelming administrative workloads, reduced staffing, and rising patient expectations. AI can help alleviate the burden of paperwork, clerical requirements, and other nonclinical tasks that have piled up over the years. 

Salesforce perspective: “AI won’t replace doctors, it will ‌augment them. People highly value and trust the personal relationship with their doctors, and AI should be seen as a tool to enhance and support that relationship, rather than a substitute. By automating administrative burdens and streamlining workflows, AI can empower doctors to deliver even better care and personalize patient experiences.” – Amit Khanna, SVP & GM, Health

By automating administrative burdens and streamlining workflows, AI can empower doctors to deliver even better care and personalize patient experiences.

Amit Khanna, SVP & GM, Health

By the numbers:

People trust physicians over AI 

• Even though nearly half of the general population internationally has used generative AI, Salesforce’s new survey found that only 2% of U.S. adults turn to generative AI sources like OpenAI’s ChatGPT or Anthropic’s Claude for healthcare information. 
• 70% turn to their physician for healthcare information — the top answer by a wide margin.
• Gen Z, however, relies on search engines the most for healthcare information. They also seek information across a wider variety of sources — for example, they’re more likely to use generative AI, social media, and health apps compared to older generations. 

“Gen Z’s reliance on multiple channels for healthcare information is one to watch due to the growing threat of misinformation. However, it also presents an opportunity for healthcare organizations to effectively market trusted information across these channels, rather than focusing on just physicians,” said Khanna.  

People are unclear on AI’s role in healthcare — but know it will have an impact

• When it comes to understanding AI’s role in healthcare, only 10% of U.S. adults say they have a good understanding of its use in patient health scenarios. The lack of understanding may affect people’s level of trust in AI in healthcare and reveals an opportunity for education.
• 47% of respondents believe AI will have a major impact on the health system overall over the next five years, but far fewer (29%) think it will have the same impact on their own patient experience.
• Lower-income respondents are even less likely to feel AI will have an impact on their experiences: 24% feel AI will have a major impact on their patient experiences, compared to 40% of high-income respondents.

People worry AI risks patient-provider relationships, security, and data accuracy

• 68% of U.S. adults are concerned that AI would weaken the relationship between patients and providers.
• 63% are concerned that AI would introduce new security risks and provide inaccurate information.
• 58% of Gen Z are particularly concerned about their data being used or exploited in unanticipated ways (e.g., data being used to determine insurance eligibility, data being used to train models). 

AI for nonclinical use cases is right in people’s comfort zone

• People are uncomfortable with AI taking on clinical tasks managed directly by doctors. Close to two-thirds of U.S. adults are uncomfortable with healthcare companies using AI to diagnose them and provide mental health support.
• People are more comfortable with AI being used to improve nonclinical experiences. For example, 69% of U.S. adults are comfortable with AI being used to schedule appointments.
• More than half are comfortable with AI being used to estimate medical expenses and simplify complex insurance information.

Today, people are uncomfortable with AI making a diagnosis. However, as the technology improves and becomes more pervasive, people could be more open to AI augmenting physicians in areas such as medical imaging, where AI models trained on vast sets of data can identity subtle patterns in images undetectable to human eyes to detect diseases, or predicting patient risk factors based on analyzing patient data.

Read more:

• Learn about Salesforce for Healthcare and Life Sciences
• Learn about Einstein Copilot: Health Actions and Salesforce’s trusted AI and data innovations in healthcare
• Find out how customers like HarmonyCares are scaling patient scheduling and Home Health with the power of Einstein 1 Platform and ‌Salesforce
• Watch how Salesforce helps healthcare organizations build unified patient profiles 
• Learn how 500 industry healthcare leaders are leveraging the power of AI to deliver equitable care, automate operations, and engage stakeholders

Methodology: Salesforce conducted a double-anonymous general population survey in partnership with YouGov in February 15-23, 2024. The total sample size was 1,443 U.S. adults with or without health insurance coverage. Results are weighted for national representation. 

Definitions: Respondents earning under $52,000 in annual household income were defined as lower income; Respondents earning between $52,000-$156,000 in annual household income were defined as middle income; Respondents earning over $156,000 in annual household income were defined as higher income. 

Gen Z are defined as being born 1997 or later; Millennials are defined as being born between 1981-1996; Gen X are defined as being born between 1965-1980; Baby Boomers are defined as being born between 1946-1964; Silent Generation are defined as being born between 1928-1945.