A new poll from health research firm KFF reveals that US adults who frequently use artificial intelligence chatbots for health advice are more likely to believe anti-vaccine myths. The survey, conducted in May with a representative sample of 2,480 US adults, found that reliance on AI tools correlates with belief in falsehoods such as vaccines causing autism or the measles vaccine being more dangerous than the virus itself. This correlation persisted even after controlling for age, race, education, and political partisanship.
Key Findings on Vaccine Misinformation
Among adults who use AI for health information at least once a week, 35% believe it is “definitely or probably true” that the measles, mumps, and rubella (MMR) vaccine causes autism. In contrast, only 20% of non-AI users hold that belief. Occasional AI users also show higher rates: 29% believe the autism myth. The false link between MMR vaccines and autism has been a cornerstone of the anti-vaccine movement, gaining traction after Robert F. Kennedy Jr. became US health secretary. The myth originated from a 1990s Lancet study that was later retracted and refuted by multiple studies.
Misinformation Across Different Myths
The poll also found that 29% of frequent AI users believe mRNA vaccines can alter DNA, compared to 20% of non-users. Additionally, 22% of frequent users think the measles vaccine is more dangerous than the measles virus, versus 15% of those who never use AI for health. These findings underscore the potential for AI to amplify vaccine misinformation.
Social Media's Role
Consulting social media for health advice similarly correlated with belief in vaccine myths. Adults who use social media for health information weekly are more than twice as likely to believe the MMR-autism myth (37% vs. 16%). The poll also noted demographic splits: lower-income and less-educated groups are more likely to seek health advice on social media, while higher-income and college-educated individuals turn to AI tools.
Context and Concerns
Concerns about AI spreading misinformation have long been raised by researchers and health officials. A March KFF survey found about a third of US adults seek health advice from AI. OpenAI acknowledged that health is a common use of ChatGPT, with hundreds of millions of health-related queries weekly. The KFF poll did not specify which AI models respondents used, noting that different chatbots vary in misinformation output due to training data and company policies. This trend mirrors search engine behavior: about 5% of Google searches are health-related, and 77% of people use search engines for new diagnoses, according to a 2025 Georgetown University study.
