Online resources for patients and the general public focused on artificial intelligence (AI) and its uses in cancer research and cancer care were found lacking in a cross-sectional analysis presented at the 2026 ASCO Annual Meeting (Abstract 9000). The study highlighted that the existing webpages and videos covering AI and cancer for the public were low-quality, difficult to read, and often failed to cover all of the risks of AI use.
“In the clinic, we hear from patients all the time, asking us about something an AI tool told them, so we know patients are using this emerging technology,” said senior study author Henry K. Litt, MD, Hematology-Oncology Fellow at the University of Pennsylvania Medicine. “Clinicians are used to educating patients about the risks of treatment, but not about the risks of misinformation that might come with using AI tools in the context of their cancer care. Our goal was to get a baseline understanding of what patients would find online if they sought out information about using AI to learn about their cancer care.”
Findings from the analysis highlight a clear need for improved resources and educational materials for patients and the public about how AI is being used in oncology.
Study Methods
The research team screened 320 webpages and videos from Google and YouTube searches for common cancer- and AI-related keywords. Once the found sites and videos were filtered for eligibility, only 52 webpages (31%) and 29 videos (19%) were included in the final analysis, with the rest excluded for not being truly patient-facing or for not being relevant.
The remaining webpages and videos were then independently evaluated by two reviewers, with discrepancies going to a third reviewer. The webpages were analyzed for readability using the validated Flesch–Kincaid, Gunning Fog, and SMOG indices. The quality of all resources was assessed using the DISCERN tool to determine consumer health information quality, with scores of 4 or higher considered high quality. Additionally, all content was evaluated for discussing key safety concepts of clinician oversight, transparency, bias, and hallucination or misinformation risk.
Key Findings
Only 33% of webpages and 23% of videos were considered high quality by the DISCERN tool.
The American Medical Association (AMA) and National Institutes of Health (NIH) recommend a reading level between sixth and eighth grade for consumer health information to maximize readability, but the median reading level for the public resources was college level.
Additionally, while more than half of the webpages and videos mentioned safety concepts of clinician oversight, transparency, and bias, only 15% mentioned the risks of hallucination or misinformation. Most of the resources also focused on how doctors and hospitals are using AI tools, rather than on how patients may be using AI in their own research and as direct users of the technology.
“While we’re always going to recommend that patients speak with their care team as their primary source of information about their care, we recognize that patients have always consulted other resources as well,” said presenting author Pearl Subramanian, MD, Internal Medicine Resident at the University of Pennsylvania Medicine. “We know that patients are going to use AI to ask questions about their cancer care, and they should have access to resources that can help them learn how to safely navigate these tools.”
“Given that just one in four items from our search were deemed relevant to patients, and among those, only one in three was high-quality, patients may struggle to find useful, plain-language information,” said co-author and study faculty advisor Ronac Mamtani, MD, who holds the David J. Vaughn MD Professorship in GU Oncology. “As AI is further integrated into oncology, patient education should be prioritized as a key part of AI implementation strategies.”
DISCLOSURE: For full disclosures of the study authors, visit coi.asco.org.
