Adults in the US who frequently seek out health advice from artificial intelligence chatbots are more likely to believe myths about vaccines, according to a poll released on Tuesday by health research firm KFF.
The survey, which was conducted in May and polled a representative sample of 2,480 US adults, found that use of AI tools and chatbots correlated with belief in falsehoods such as vaccines causing autism or that the measles vaccine poses more danger than the corresponding virus. The connection remained while controlling for factors such as age, race, education and political partisanship.
Concern over how AI may spread misinformation and influence public opinion has long been an issue among researchers and health officials. A large percentage of Americans have begun turning to AI chatbots for medical advice, repeated polls show, with another KFF survey from March finding that about a third of US adults seek out health advice from AI.
AI firms have also acknowledged the prevalence of queries about medical matters. “Health is already one of the most common ways people use ChatGPT, with hundreds of millions of people asking health and wellness questions each week,” OpenAI said in a January blog post announcing the creation of a specialized ChatGPT Health tool.
Among US adults who use AI tools to find health information at least once a week, KFF’s poll found that 35% of them believe that it is “definitely or probably true” that measles, mumps and rubella (MMR) vaccines have been proven to cause autism in children. Only 20% of US adults who do not use AI for health hold a similar belief, while 29% of US adults who occasionally consult AI for health believe that myth, according to the poll.
The falsehood that MMR vaccines cause autism is a key pillar of the anti-vaccine movement, which has gained additional influence after the Covid-19 pandemic and appointment of Robert F Kennedy Jr as the US health secretary. Kennedy and others associated with the anti-vaccine movement have long used debunked or retracted medical studies to advance their views. The myth that MMR vaccines bring on autism gained prominence after the Lancet journal published a study in the 1990s which was later fully retracted after its findings were found to be false. It has since been refuted by multiple other studies.
The KFF poll additionally found that 29% of US adults who frequently use AI tools for health believe that mRNA vaccines can change your DNA, which is not true, whereas only 20% of people who never use AI hold that belief. Among frequent AI users, 22% believe that the measles vaccine is more dangerous than the measles virus – compared with only 15% of people who do not use AI for health.
Consulting social media platforms for health advice also correlated with belief in misinformation around vaccines in the poll’s results. KFF found “adults who use social media for health information at least weekly are more than twice as likely as those who don’t use social media for health to say the myth linking MMR vaccines to autism is ‘probably’ or ‘definitely true’ (37% v. 16%)”.
The poll also found a split between which demographics sometimes look to social media for health advice versus who consults AI. Lower income groups and people with less than college education are more likely to seek out advice via social media, while a higher percentage of people in households making above $90,000 per year or who possess a college education turn to AI tools.
skip past newsletter promotion
Sign up to TechScape
A weekly dive in to how technology is shaping our lives
after newsletter promotion
The KFF poll did not ask which AI models respondents used while seeking out health advice. Different chatbots produce varying levels of misinformation, and each contains its own biases as a result of training data and respective companies’ decisions on how the bots should respond to divisive questions.
Using chatbots to find health information continues a longstanding pattern in how people use search engines: about 5% of all Google searches concern health, and about 77% of people use search engines to ask about new diagnoses, according to a 2025 research paper by a Georgetown University researcher.
