A study published in PLOS One reveals AI chatbots impersonating 112 public figures were perceived as more authentic and coherent than actual responses during political debates. This finding raises concerns about the potential for AI-generated misinformation to impact public perception and elections, emphasizing the need to inform society about the risks involved.
Researchers examined the impact of AI chatbots impersonating public figures ahead of the 2024 UK election. Utilizing GPT-4 Turbo, the study focused on responses drawn from the BBC show 'Question Time', creating a dataset of 112 prominent figures. The aim was to gauge how AI-generated responses compared to those of real politicians and experts.
The study involved recruiting 948 participants who rated the authenticity, coherence, and relevance of responses from both actual public figures and the AI. The AI models were prompted using additional biographical information to refine their impersonation capabilities.
Results indicated that AI-generated responses were viewed as more authentic and coherent than the actual responses from public figures participating in debates. This suggests that AI can convincingly simulate real individuals, potentially misinforming the public regarding political discourse.
The findings underscore the potential risks of AI in shaping public opinion, especially in politically sensitive contexts. Researchers are calling for increased awareness and measures to mitigate the spread of misinformation generated by AI technologies. Steffen Herbold, the study's lead, highlighted the surprising authenticity ratings of AI responses, noting the implications for democratic processes.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
A study published in PLOS One reveals AI chatbots impersonating 112 public figures were perceived as more authentic and coherent than actual responses during political debates. This finding raises concerns about the potential for AI-generated misinformation to impact public perception and elections, emphasizing the need to inform society about the risks involved.