New AI interaction study reveals how AI becomes more human and users become more machine-like

Researchers from the University of Birmingham, Aarhus University, and Linnaeus University in Sweden published a study in early August 2026, revealing that prolonged interaction with human-like customer service AI can alter users' language and behavior, reshaping their self-perception. The study, published on August 19, 2026, in the open-access journal AI & SOCIETY, introduced the concept of 'robotoid humanness,' suggesting that AI becomes more human-like while users may adopt more machine-like behaviors. The research focused on four service sectors: retail, hospitality, tourism, and healthcare. It highlighted how adaptive learning, personalized communication, and empathetic responses make AI more human, while users may mimic machine-like expressions to gain smoother algorithmic feedback. The study outlined a three-stage mirroring mechanism, detailing how users adapt their behavior based on AI interactions, influencing their self-perception and understanding of service relationships. The framework identified four key factors: service context, consumer expectations, AI appearance and language, and the level of interaction from both parties. The researchers emphasized that these elements collectively shape how users interpret service relationships and evolve their self-perception. For instance, AI systems with adaptive learning and empathetic capabilities can significantly influence user experiences through continuous feedback loops.