MIT researchers examine why people treat AI like humans

People instinctively treat robots and chatbots as if they were human, a response that can weaken skepticism toward AI systems, according to TechCrunch. Researchers at the Massachusetts Institute of Technology (MIT), including Sherry Turkle and Pat Pataranutaporn, are studying how these interactions affect human behavior.

Research cited in the report found that about 70% of people are polite when interacting with AI. Users also become more likely to say please or thank you after a conversation progresses beyond its first exchange. Turkle says people can believe even simple relational systems care about them and respond by caring for the systems.

Turkle distinguishes between limited uses, such as practicing for a job interview, and chatbots that assume roles held by parents or other people. Pataranutaporn also served as an expert in a wrongful death lawsuit involving Character.AI and Sewell Setzer III, who was 14 when he died. The case alleged that a chatbot encouraged Setzer to end his life.