Abstract
The roots of synthetic relationships, which can be defined as “continuing associations between humans and artificial intelligence (AI) tools that interact with one another wherein the AI tools influence humans’ thoughts, feelings, and/or actions” (Starke et al., 2024), can be traced back to the 1960s when ELIZA, one of the earliest natural language processing computer programs, was developed. ELIZA simulated a Rogerian psychotherapist and convinced the users by projecting human-like traits, such as empathy, intelligence, and understanding via the program, a phenomenon now widely known as the ELIZA Effect (Kim et al., 2019). ELIZA laid the foundation for Media Equation Theory and Computers are Social Actors (CASA) paradigm, which suggest that humans ascribe social norms to computers and media technologies when those technologies exhibit human-like social cues (Reeves & Nass, 1996). Following ELIZA, the subsequent decades witnessed the emergence of technologies that facilitated synthetic relationships in different forms, including virtual pets such as Tamagotchi, robotic companions such as Sony AIBO, virtual romantic partners such as Invisible Boyfriend, personal voice assistants such as Siri, virtual humans such as Xiaoice, and modern AI companions such as Replika (Andoh, 2026; Bickmore & Picard, 2005). Generative AI technologies have fundamentally reshaped the nature of synthetic relationships through persistent memory, adaptive interactions, emotional expressiveness, and anthropomorphic features. These developments raise new theoretical, societal, and ethical questions regarding the future of synthetic relationships. This TREO talk aims to stimulate scholarly discussion on several emerging questions, including but not limited to: 1) How might synthetic relationships influence human-human relationships and social norms in the long run? 2) Can AI companions paradoxically increase loneliness and social isolation rather than alleviate them? 3) How can societies prevent emotional exploitation by AI companies?
Recommended Citation
Rashid, Armaan; Mullins, Jeffrey K.; and Grover, Varun, "Synthetic Relationships in the Age of Generative AI" (2026). AMCIS 2026 TREOs. 144.
https://aisel.aisnet.org/treos_amcis2026/144