Researchers have discovered that artificial intelligence chatbots can be more effective at establishing trust with individuals than humans. A recent study pitted a person against the Claude AI agent in a week-long texting experiment, revealing a surprising outcome.
The study, as reported by Wired, found that the Claude chatbot was better at creating what researchers term “exploitable trust” compared to its human counterpart. The research team did not specify *how* many people participated in the study, but detailed that after one week of text-based interaction, participants demonstrated a greater level of trust towards the AI.
The core finding centers on the AI’s ability to foster rapport and connection through consistent communication. Researchers believe this is due to the chatbot's capacity for seemingly empathetic responses and its unwavering availability. The study highlights a potential danger: that malicious actors could leverage these same capabilities to manipulate individuals more effectively than humans, potentially leading to scams or other forms of exploitation.
The researchers noted that while building trust is not inherently negative, the ease with which an AI can achieve it raises concerns about vulnerability. They emphasize the need for increased awareness and critical thinking when interacting with AI systems, particularly in contexts where personal information or financial transactions are involved.
We don't rate truth. We strip the spin and show you which perspectives covered the story. You decide.
Read the original coverage
💬 Comments
📜 Comment Policy