AI vs. Pseudoscience: Finland Develops Chatbots to Help Identify Health Misinformation

The rapid development of generative artificial intelligence (AI) has made it easier for false and misleading health information to spread on a large scale. However, researchers are now attempting to turn AI into a new tool to combat misinformation by simulating the "cognitive vaccination" process, helping people enhance their ability to identify pseudoscience and health rumors.

Xinhua News Agency reported that researchers from the University of Oulu in Finland, in collaboration with international colleagues, have developed an AI chatbot. By using the "cognitive vaccination" method, it can help people resist misinformation in the field of health.

Oulu University recently released a press release stating that in recent years, generative AI tools have developed rapidly, making it easier for false or misleading content to be mass-produced and disseminated. At the same time, AI can also be used to combat misinformation.

For instance, this chatbot named Forty employs the "cognitive inoculation" approach derived from social psychology. The basic idea is to have people exposed to "reduced versions" of incorrect viewpoints under controlled conditions, thereby enhancing their resistance to false information in real-life situations.

The research team said that this principle is similar to how the human body enhances its immune system through vaccination.

The research team deployed Forty on a public website, conducting "structured" conversations with users on common health topics such as daily tooth brushing, alcohol consumption, physical activity and mental health. This refers to the chatbot guiding users through a pre-set scientific education process, helping them identify misleading viewpoints, analyze logical flaws, and form a more solid and accurate understanding.

To verify the effectiveness of the system, the researchers invited 65 participants to have a conversation with the chatbot, Forty. The results showed that compared to simply reading or writing, interacting with this chatbot was more effective in enhancing participants' resistance to incorrect health information.

The researchers stated that their next plan is to expand the application scope of chatbots to more complex life situations, and explore how conversational AI can assist people in dealing with daily stress, unexpected changes and other challenges.