Concerns surrounding the use of artificial intelligence (AI) in healthcare are widespread among Americans, particularly regarding privacy and data security risks associated with open-source AI models. Despite this unease, many individuals are increasingly turning to AI chatbots for medical guidance.
Recent findings from the Pew Research Center reveal that approximately one-third of American adults have utilized AI chatbots for health-related diagnoses and treatments, with a significant majority finding these services beneficial.
As governments worldwide grapple with the implications of generative AI, the technology has become an integral part of everyday life. While critics highlight its potential for misuse, proponents argue that AI has the potential to transform healthcare delivery—though opinions are divided on whether these changes are positive or negative.
The latest Pew data indicates that while some users have replaced traditional healthcare consultations with AI tools, a considerable number of people express concern regarding the extent of AI's integration into medical services. The survey noted that over 70% of respondents feel it is crucial for their healthcare providers to disclose any AI usage during their treatment.
In contrast, only 22% of participants reported a clear understanding of how AI technology is implemented in their healthcare, with nearly half uncertain about its use. More than 50% of respondents wish they had a greater say in whether AI is employed in their medical care.
The public’s awareness of AI surged with the launch of ChatGPT in 2022, but within healthcare, AI encompasses a far broader array of technologies. Tim Curry, a public health lecturer, emphasized that his research into machine learning applications in medicine has been ongoing for decades. He highlighted a disconnect between public perception and the realities of AI's role in healthcare.
Despite the lack of trust in AI among some demographics, a substantial percentage of people—over 80%—want to be informed about AI's role in the interpretation of medical scans and diagnoses. Jyoti Pathak, dean of Arizona State University’s School of Technology for Public Health, underscored the necessity of clear communication, stating that patient trust is fundamental to the widespread acceptance of AI in healthcare settings.
AI’s applications in diagnostics pose the greatest risks, according to Pathak, necessitating rigorous FDA regulation, clinical validation, and transparency. In contrast, administrative AI used for tasks like scheduling carries lower risks but still requires patient education and consent.
Despite skepticism, surveys indicate that many individuals are opting for AI chatbots over professional medical consultations. More than one-third of over 4,000 surveyed Americans reported using AI chatbots for medical inquiries, often citing affordability as a motivating factor.
Younger adults, particularly those aged 18-29, show the highest engagement with AI chatbots, despite their general apprehension toward AI. This demographic is significantly affected by rising healthcare costs and lower insurance coverage compared to other age groups.
However, relying on AI chatbots can be problematic, as many services require users to agree to terms that allow their interactions to contribute to future AI training, potentially compromising personal health data. Moreover, the current development stage of these chatbots may not guarantee reliable medical advice, leading to misinformation.
Pathak cautioned that without access to appropriate medical histories, AI chatbots could mislead users, instilling either unwarranted optimism or unnecessary alarm. Curry proposed that medical practices might soon create their own AI systems to handle health-related inquiries, which would be HIPAA compliant, unlike existing open-source models.
Until such solutions are available, Curry advises patients who cannot visit a doctor to consult credible sources like Mayo Clinic and Cleveland Clinic for peer-reviewed health information.
