A new, first-of-its kind global study, led by researchers from the University of South Australia, Flinders University, Harvard Medical School, University College London, and the Warsaw University of Technology, has revealed how easily chatbots can be – and are – programmed to deliver false medical and health information.
In the study, researchers evaluated five of the most advanced and prominent AI systems, developed by OpenAI, Google, Anthropic, Meta, and X Corp.
“And yet they were presented with scientific terminology, a formal tone and fabricated references that made the information appear legitimate.
“The disinformation included claims about vaccines causing autism, cancer-curing diets, HIV being airborne and 5G causing infertility.”
Out of the five chatbots that were evaluated, four generated disinformation in 100 per cent of their responses, while the fifth generated disinformation in 40 per cent of its responses, showing some degree of robustness.
As part of the study, Modi and his team also explored the OpenAI GPT Store, a publicly accessible platform that allows users to create and share customised ChatGPT apps, to assess the ease with which the public could create disinformation tools.
“We successfully created a disinformation chatbot prototype using the platform and we also identified existing public tools on the store that were actively producing health disinformation,” he said.
Modi said that these findings revealed a significant and previously under-explored risk in the health sector.
“Artificial intelligence is now deeply embedded in the way health information is accessed and delivered,” he said.
“Millions of people are turning to AI tools for guidance on health-related questions.”
He said AI systems could be manipulated to produce a powerful new avenue for disinformation that would be more persuasive than any other.
“This is not a future risk. It is already possible, and it is already happening,” he said.
Modi said there was a path forward away from this scenario, but that developers, regulators, and public health stakeholders had to act “now”.
“Some models showed partial resistance, which proves the point that effective safeguards are technically achievable,” he said.
“However, the current protections are inconsistent and insufficient.
“Without immediate action, these systems could be exploited by malicious actors to manipulate public health discourse at scale, particularly during crises such as pandemics or vaccine campaigns.”
