The study, published in PNAS Nexus, investigated whether labeling content as AI-generated changes people’s reactions to political and public policy messages. Researchers surveyed 1,601 Americans and presented them with AI-generated arguments on a variety of policy issues, including college athlete pay, drug importation, geoengineering, and the liability of social media platforms. Participants were randomly told that the message was created by an AI systems expert or a human policy expert, or they were given no information about the source.
They found that messages were generally persuasive, regardless of the label given to them. On average, participants’ support for a policy changed by nearly 10 percentage points after reading the discussion. However, the researchers found no significant difference between the effects of AI-labeled messages, human-labeled messages, and unlabeled content.
Notably, most participants accepted the author name label as accurate. More than 90% of people who saw the AI or human label believed the information about the message’s author. However, this awareness did not significantly influence whether they changed their views, considered the information accurate, or intended to share it with others.
The findings come as governments around the world consider rules requiring AI-generated content to be clearly identified. The European Union’s AI law and several U.S. legislative proposals include provisions aimed at increasing transparency through labeling. The findings suggest that while such measures may help people understand the source of content, they are unlikely to reduce the impact of AI-generated information on their own, the researchers said.
The study also found that the lack of influence was largely consistent across political party affiliation, education level, age, familiarity with AI, and prior knowledge on the topic. Older participants showed slightly more skepticism toward AI-labeled content, but the overall effect remained small.
Previously, Qazinform news agency reported that an increasing number of children are exposed to AI-generated videos on TikTok.
