A new poll released Tuesday by health‑policy research firm KFF finds that U.S. adults who regularly turn to artificial‑intelligence chatbots for medical information are significantly more likely to endorse false claims about vaccines, including the long‑discredited idea that shots cause autism. The correlation persists after accounting for age, race, education and political affiliation, raising fresh concerns about the role of AI‑generated advice in public‑health misinformation.
What happened
The online survey was fielded in May 2026 and sampled a nationally representative group of 2,480 U.S. adults. Respondents were asked how frequently they used AI tools such as ChatGPT, Google Gemini or Bard for health‑related queries. Roughly 28 % said they consulted an AI chatbot “often” or “very often” for medical information. Among this frequent‑user cohort, 42 % agreed with at least one vaccine‑related myth, compared with 24 % of those who rarely or never used AI for health advice.
The most commonly endorsed falsehoods were:
* COVID‑19 vaccines alter DNA – 31 % of frequent AI users
* Vaccines cause autism – 27 % of frequent AI users (versus 12 % of infrequent users)
* The measles vaccine is more dangerous than the disease it prevents – belief rates not broken out by usage level but included in the overall myth endorsement figure.
Overall, 68 % of all respondents reported trusting AI chatbots “somewhat” or “a lot” for general health questions, despite platform disclaimers that the tools are not substitutes for professional medical advice. Moreover, 54 % of frequent AI users said they had not verified the chatbot’s answers with a healthcare professional.
Why it matters
Vaccination rates in the United States have plateaued, and public‑health officials are seeking to boost uptake of newer respiratory vaccines. Misperceptions about vaccine safety can erode confidence and hinder campaign goals. The KFF poll suggests that AI chatbots—now embedded in consumer search habits—may be an under‑examined vector for spreading vaccine misinformation, complementing the well‑documented role of social‑media platforms.
If AI tools are reinforcing or amplifying anti‑vaccine narratives, the impact could extend beyond individual health decisions to broader public‑health outcomes, including potential outbreaks of preventable diseases. The findings also intersect with ongoing regulatory discussions about the accountability of AI‑driven health applications, which currently operate under voluntary disclaimer policies rather than enforceable standards.
Background and context
AI chatbots have surged in popularity since the release of large language models in 2023, with major tech firms integrating them into search engines, messaging apps and dedicated health portals. While these models can synthesize information quickly, they draw on training data that includes a mix of peer‑reviewed literature, news articles, forums and other internet content—some of which may be inaccurate or outdated.
KFF’s survey is the first large‑scale, nationally representative poll to link AI health‑information usage with vaccine myth endorsement. Earlier research has documented that exposure to misinformation on social media correlates with vaccine hesitancy, but the role of AI‑generated content has remained largely speculative. The firm controlled for demographic and partisan variables, indicating that the observed association is not simply a reflection of known predictors of vaccine skepticism.
Competing claims and uncertainty
KFF’s senior researcher, Dr. Maya Patel, cautioned that the poll shows correlation, not causation. “It is possible that individuals already skeptical of mainstream medicine are drawn to AI tools that appear to validate their doubts,” she said. The firm’s methodology does not establish whether chatbot use actively creates belief in myths, merely that the two variables move together.
Industry representatives for AI platforms have not yet responded to the poll’s release. Some analysts argue that the responsibility for misinformation lies primarily with users who fail to verify information, rather than with the underlying technology. Others point to the lack of real‑time fact‑checking mechanisms within current chatbot architectures as a structural weakness that could be addressed through tighter oversight or mandatory accuracy reporting.
Public‑health experts emphasize that the 54 % of frequent AI users who did not seek professional verification represent a vulnerable subgroup. Yet the poll does not capture the depth of users’ engagement—whether they rely on a single chatbot response or cross‑check multiple sources—leaving open questions about how misinformation spreads within AI interactions.
What to watch next
Regulators at the Federal Trade Commission and the Food and Drug Administration have signaled interest in developing guidelines for AI health tools, including potential labeling requirements that more prominently flag non‑clinical content. Legislative proposals introduced in the 118th Congress could mandate independent audits of AI models used for medical advice.
Consumer‑advocacy groups are likely to launch campaigns urging users to treat AI chatbots as supplemental, not primary, sources of health information. Monitoring future KFF surveys or similar studies will be essential to gauge whether the observed correlation strengthens, weakens, or shifts as AI adoption evolves.
Additionally, tech companies may roll out built‑in verification features, such as linking chatbot answers to vetted medical databases or prompting users to consult a licensed professional. Tracking the uptake and effectiveness of such safeguards will be a key indicator of industry response to the poll’s findings.
Conclusion
The KFF poll provides the first robust evidence that frequent reliance on AI chatbots for health advice is associated with higher acceptance of vaccine myths among U.S. adults. While the data do not prove that chatbots cause misinformation, the correlation persists after adjusting for known demographic and partisan factors, suggesting that AI tools could be an overlooked conduit for public‑health risk. As AI becomes increasingly woven into everyday information seeking, policymakers, tech firms and health officials will need to address the accuracy of AI‑generated medical content and promote verification practices to safeguard vaccination efforts.
Sources
– The Guardian, “Frequent AI chatbot users more likely to believe anti‑vaccine myths, poll finds,” June 30 2026, https://www.theguardian.com/technology/2026/jun/30/ai-chatbot-use-anti-vaccine-myths-poll
Story synopsis gathered from: The Guardian World — source
Corrections
If you believe this article contains an error, contact Herald Express with the source URL and supporting evidence.
