Patient-First Health with Generative AI: Reshaping the Care Experience

The report explores how generative AI can be used as productivity boosters, insights generators, and action drivers to empower patients and reduce health system burdens. It identifies key use cases in health education, patient triage, and disease management, while highlighting barriers such as data inaccuracy, bias, and scalability in low-resource environments.

• Generative AI use cases for patient engagement are categorized into three types: productivity boosters that automate manual tasks and extract insights from unstructured data; insights generators that act as 'co-pilots' for real-time analysis of structured and unstructured datasets; and action drivers that provide intuitive, human-like conversations to inform care decisions.
• Generative AI can address critical gaps in health education by providing reliable, polylingual information that is accessible regardless of a user's education level or cultural context, potentially reducing the economic costs of low health literacy, which are estimated at $238 billion annually in the US and between €120 billion and €320 billion in the EU.
• The technology can alleviate healthcare provider burnout and system strain by serving as co-pilots for patient triage and intake, helping to stratify patients and route them to appropriate care levels, particularly in regions with severe worker shortages.
• In disease management, generative AI can improve treatment adherence—which is currently missed by as many as 60% of chronic illness patients in their first year—by creating personalized interventions based on predictive algorithm outputs.
• The primary barriers to adopting patient-facing generative AI include mistrust due to 'hallucinations' (incorrect or misleading answers), biased outputs reflecting training data, patient privacy concerns, and a lack of high-quality training data for less common languages.
• Scalability in low- and middle-income countries (LMICs) is hindered by the high cost of collecting and storing large volumes of data and the heavy computational and cloud computing resources required to run large models.
• To successfully implement generative AI, stakeholders should prioritize fine-tuning models for empathy, connecting disconnected data layers to mitigate bias, maintaining 'humans in the loop' to review AI recommendations, and developing cost-efficient deployment models for diverse regional contexts.

APA

World Economic Forum (2024). Patient-First Health with Generative AI: Reshaping the Care Experience. https://www3.weforum.org/docs/WEF_Patient_First_Health_with_Generative_AI_2024.pdf

Chicago

World Economic Forum. Patient-First Health with Generative AI: Reshaping the Care Experience. 2024. https://www3.weforum.org/docs/WEF_Patient_First_Health_with_Generative_AI_2024.pdf.

Wikipedia

{{cite report |author=World Economic Forum |title=Patient-First Health with Generative AI: Reshaping the Care Experience |date=January 2024 |url=https://www3.weforum.org/docs/WEF_Patient_First_Health_with_Generative_AI_2024.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{worldeconomicforum2024patientfirst, author = {{World Economic Forum}}, title = {{Patient-First Health with Generative AI: Reshaping the Care Experience}}, institution = {World Economic Forum}, year = {2024}, month = jan, url = {https://www3.weforum.org/docs/WEF_Patient_First_Health_with_Generative_AI_2024.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }