This research investigates how different demographic groups interact with and are influenced by misinformation when engaging with large language models.

Key findings:

• Demographic factors significantly impact susceptibility to LLM-generated misinformation
• Bidirectional persuasion exists between humans and LLMs, with each influencing the other
• Certain population groups demonstrate higher vulnerability to AI-generated misinformation
• Understanding these dynamics is crucial for developing targeted protective measures

Security implications: This research highlights critical vulnerabilities in our information ecosystem, providing insights that can help develop demographic-aware safeguards against AI-amplified misinformation campaigns.

Persuasion at Play: Understanding Misinformation Dynamics in Demographic-Aware Human-LLM Interactions